This study delves into hospital mobility, understood as an indicator of perceived service quality, across the Italian regions of Apulia and Emilia Romagna, utilizing logistic regression among machine learning techniques. The focus is on how structural, operational, and clinical variables impact patient perceptions of service quality, influencing their healthcare choices. Through the analysis of mobility trends with machine learning, significant differences between regions were uncovered, highlighting the influence of regional context on perceived quality. The integration of SHAP (SHapley Additive exPlanations) values into our analysis provided deeper insights into the logistic regression model, elucidating the specific contribution of each variable to perceived healthcare quality. This incorporation of SHAP values underscores the study's commitment to employing advanced, explainable AI techniques to enhance the interpretability and fairness of healthcare service evaluations. The choice of logistic regression elucidated the impact of specific variables on quality perception, offering essential insights for optimizing healthcare resource distribution and underscoring the importance of data-driven strategies to foster more equitable, efficient, and patient-centred healthcare systems. Contributing to the understanding of perceived quality dynamics within the healthcare context, the research paves the way for further investigations into enhancing accessibility and service quality, leveraging machine learning as a tool for improving healthcare services efficiency in diverse regional settings.

This study addressed the need for efficient data collection in micro-mobility using low-end hardware, focusing on developing a system capable of handling large real-time data streams. We proposed a novel data collection system that leverages incremental learning techniques, enabling AI-based services for micro-mobility passengers through smartphones. Our approach utilizes 3-axis acceleration sensors integrated into micro-mobility devices, collecting motion data in real-time. This data is processed through an incremental learning pre-processing server, employing the Broad Learning System algorithm, to ensure robust performance in noisy, real-world environments. The use of incremental learning is pivotal, allowing our system to adapt continuously to new data, enhancing the accuracy and relevance of the AI services provided to passengers. This research not only advances the technical capabilities in micro-mobility data handling but also has the potential to enhance the overall passenger experience through improved AI-based services.

Barrett's esopaghagus (BE) is a known precursor of esophageal adenocarcinoma (EAC). Patients with BE undergo regular surveillance to early detect stages of EAC. Volumetric laser endomicroscopy (VLE) is a novel technology capable of imaging the inner tissue layers of the esophagus over a 6-cm length scan. However, interpretation of full VLE scans is still a challenge for human observers. In this work, we train an ensemble of deep convolutional neural networks to detect neoplasia in BE patients, using a dataset of images acquired with VLE in a multicenter study. We achieve values of AUC=$0.96$ on the unseen test dataset and we compare our results with previous work done with VLE analysis. Our method for detecting neoplasia in BE patients facilitates future advances on patient treatment and provides clinicians with new assisting solutions to process and better understand VLE data.

Mobility is a vital welfare indicator which may influence broilers’ daily activities. Classical broiler mobility assessment methods are laborious and cannot provide timely insights into their conditions. Here, we proposed a semi-supervised Deep Learning (DL) model, YOLOv5, combined with Deep Sort algorithm conjoined with our newly proposed algorithm, Neo-Deep Sort, for individual broiler mobility tracking. Initially, 1,650 labeled images from five days were employed to train the YOLOv5 model. Through semi-supervised learning (SSL), this narrowly trained model was then used for pseudo-labeling 2,160 images, of which 2,153 were successfully labeled. Thereafter, the YOLOv5 model was fine-tuned on the newly labeled images. Lastly, the trained YOLOv5 and the Neo-Deep Sort algorithm were applied to detect and track 28 broilers in two pens and categorized them in terms of hourly and daily traveled distances and speeds. SSL helped in increasing the YOLOv5 model’s mean Average Precision (mAP), in detecting birds, from 81% to 98%. As compared with the manually measured covered distances of broilers, the combined model provided individual broiler's hourly moved distances with a validation accuracy of about 80%. Eventually, individual and flock level mobilities were quantified while overcoming the occlusion, false and miss detection issues.

The ongoing COVID-19 global pandemic is affecting every facet of human lives (e.g., public health, education, economy, transportation, and the environment). This novel pandemic and citywide implemented lockdown measures are affecting virus transmission, people’s travel patterns, and air quality. Many studies have been conducted to predict the COVID-19 diffusion, assess the impacts of the pandemic on human mobility and air quality, and assess the impacts of lockdown measures on viral spread with a range of Machine Learning (ML) techniques. This review study aims to analyze results from past research to understand the interactions among the COVID-19 pandemic, lockdown measures, human mobility, and air quality. The critical review of prior studies indicates that urban form, people's socioeconomic and physical conditions, social cohesion, and social distancing measures significantly affect human mobility and COVID-19 transmission. During the COVID-19 pandemic, many people are inclined to use private transportation for necessary travel purposes to mitigate coronavirus-related health problems. This review study also noticed that COVID-19 related lockdown measures significantly improve air quality by reducing the concentration of air pollutants, which in turn improves the COVID-19 situation by reducing respiratory-related sickness and deaths of the people. It is argued that ML is a powerful, effective, and robust analytic paradigm to handle complex and wicked problems such as a global pandemic. This study also discusses policy implications, which will be helpful for policymakers to take prompt actions to moderate the severity of the pandemic and improve urban environments by adopting data-driven analytic methods.

Recently, smart mobility intelligent traffic services have become a critical task in Intelligent Transportation Systems (ITS). This involves not only the use of advanced sensors and controllers but also the ability to respond to real-time traffic situations at intersections, alleviate congestion, and generate policies to prevent traffic jams. DRL (Deep Reinforcement Learning) provides a natural framework for processing tasks. In DRL, each intersection can control itself and coordinate with neighbors to achieve optimal network-wide policies. However, comparing approaches remains a challenging task due to the existence of numerous possible configurations. This research performs a critical comparison of various traffic controllers found in the literature. It demonstrates that using a nonlinear approximator for coordination mechanisms and enhancing observability at each intersection are key performance drivers.

Over a billion people around the world are disabled, among them, 253 million are visually impaired or blind, and this number is greatly increasing due to ageing, chronic diseases, poor environment, and health. Despite many proposals, the current devices and systems lack maturity and do not completely fulfill user requirements and satisfaction. Increased research activity in this field is required to encourage the development, commercialization, and widespread acceptance of low-cost and affordable assistive technologies for visual impairment and other disabilities. This paper proposes a novel approach using a LiDAR with a servo motor and an ultrasonic sensor to collect data and predict objects using deep learning for environment perception and navigation. We adopted this approach in a pair of smart glasses, called LidSonic V2.0, to enable the identification of obstacles for the visually impaired. The LidSonic system consists of an Arduino Uno edge computing device integrated into the smart glasses and a smartphone app that transmits data via Bluetooth. Arduino gathers data, operates the sensors on smart glasses, detects obstacles using simple data processing, and provides buzzer feedback to visually impaired users. The smartphone application collects data from Arduino, detects and classifies items in the spatial environment, and gives spoken feedback to the user on the detected objects. In comparison to image processing-based glasses, LidSonic uses far less processing time and energy to classify obstacles using simple LiDAR data, according to several integer measurements. We comprehensively describe the proposed system's hardware and software design, construct their prototype implementations, and test them in real-world environments. Using the open platforms, WEKA and TensorFlow, the entire LidSonic system is built with affordable off-the-shelf sensors and a microcontroller board costing less than $80. Essentially, we provide designs of an inexpensive, miniature, green device that can be built into, or mounted on, any pair of glasses or even a wheelchair to help the visually impaired. Our approach affords faster inference and decision-making using relatively low energy with smaller data sizes as well as faster communications for the edge, fog, and cloud computing.

The Norwegian Museum of Science and Technology have developed a learning concept for school classes in science centres named ‘learning trails’. In this concept, groups of students perform a series of thematically related experiments with installations in the science centre. The learning trails are designed to support the generic learning outcomes for science centre visits. We argue for using the previously developed Engagement Profile in an indicator to determine both media forms and generic learning outcomes for such learning concepts. Further, we implemented the learning trails in two modes: one mode used paper-based content to guide the students, while the other mode supported the use of tablet PCs where engaging content is triggered when the students approach the location of an experiment in the learning trail. We studied the engagement factors of the learning trails and observed how school classes use these. In a study with 113 students from lower secondary school, they answered short questionnaires that were integrated into the implementation of the learning trails. While the concept of the learning trails was evaluated positively, we could not find significant differences in how engaging the two implemented modes were.

Virtual screening - predicting which compounds within a specified compound library bind to a target molecule, typically a protein - is a fundamental task in the field of drug discovery. Doing virtual screening well provides tangible practical benefits, including reduced drug development costs, faster time to therapeutic viability, and fewer unforeseen side effects. As with most applied computational tasks, the algorithms currently used to perform virtual screening feature inherent tradeoffs between speed and accuracy. Furthermore, even theoretically rigorous, computationally intensive methods may fail to account for important effects relevant to whether a given compound will ultimately be usable as a drug. Here we investigate the virtual screening performance of the recently released Gnina molecular docking software, which uses deep convolutional networks to score protein-ligand structures. We find, on average, that Gnina outperforms conventional empirical scoring. The default scoring in Gnina outperforms the empirical AutoDock Vina scoring function on 89 of the 117 targets of the DUD-E and LIT-PCBA virtual screening benchmarks with a median 1% early enrichment factor that is more than twice that of Vina. However, we also find that issues of bias linger in these sets, even when not used directly to train models, and this bias obfuscates to what extent machine learning models are achieving their performance through a sophisticated interpretation of molecular interactions versus fitting to non-informative simplistic property distributions.

There is only a very short reaction time for people to find the best way out of a building in a fire outbreak. Software applications can be used to assist the rapid evacuation of victims; however, this is an arduous task, which requires an understanding of advanced technologies. Since well-known pathway algorithms (such as, Dijkstra, Bellman-Ford and A*) can lead to serious problems over performance, we decided to make use of deep reinforcement learning techniques. A wide range of strategies including a random initialization of replay buffer and transfer learning were assessed in three projects involving schools of different sizes. The results showed the proposal was viable and that in most cases the performance of transfer learning was superior. In addition, the study raised challenges that had to be faced in the future.

Deep learning has achieved lots of successes in many fields, but when trainable sample are extremely limited, deep learning often under or overfitting to few samples. Meta-learning was proposed to solve difficulties in few-shot learning and fast adaptive areas. Meta-learner learns to remember some common knowledge by training on large scale tasks sampled from a certain data distribution to equip generalization when facing unseen new tasks. Due to the limitation of samples, most approaches only use shallow neural network to avoid overfitting and reduce the difficulty of training process, that causes the waste of many extra information when adapting to unseen tasks. Euclidean space-based gradient descent also make meta-learner's update inaccurate. These issues cause many meta-learning model hard to extract feature from samples and update network parameters. In this paper, we propose a novel method by using multi-stage joint training approach to post the bottleneck during adapting process. To accelerate adapt procedure, we also constraint network to Stiefel manifold, thus meta-learner could perform more stable gradient descent in limited steps. Experiment on mini-ImageNet shows that our method reaches better accuracy under 5-way 1-shot and 5-way 5-shot conditions.

Deep learning has been widely used in various tasks such as computer vision, natural language processing, and predictive analysis, recommendation systems in the past decade. However, practical scenarios often lack labeled data, posing challenges for traditional supervised methods. Semi-supervised classification methods address this by leveraging both labeled and unlabeled data to enhance model performance, but they face challenges in effectively utilizing unlabeled data and distinguishing reliable information from unreliable sources. This paper introduces ReliaMatch, a semi-supervised classification method that addresses these challenges by using a confidence threshold. It incorporates a curriculum learning stage, feature filtering, and pseudo-label filtering to improve classification accuracy and reliability. The feature filtering module eliminates ambiguous semantic features by comparing labeled and unlabeled data in the feature space. The pseudo-label filtering module removes unreliable pseudo-labels with low confidence, enhancing algorithm reliability. ReliaMatch employs a curriculum learning training mode, gradually increasing training dataset difficulty by combining selected samples and pseudo-labels with labeled data. This supervised approach enhances classification performance. Experimental results show that ReliaMatch effectively overcomes challenges associated with the underutilization of unlabeled data and the introduction of error information, outperforming the pseudo-label strategy in semi-supervised classification.

The increasing trend of studying the innate softness of robotic structures and amalgamating it with the benefits of the extensive developments in the field of embodied intelligence has led to sprouting of a relatively new yet extremely rewarding sphere of technology. The fusion of current deep reinforcement algorithms with physical advantages of a soft bio-inspired structure certainly directs us to a fruitful prospect of designing completely self-sufficient agents that are capable of learning from observations collected from their environment to achieve a task they have been assigned. For soft robotics structure possessing countless degrees of freedom, it is often not easy (something not even possible) to formulate mathematical constraints necessary for training a deep reinforcement learning (DRL) agent for the task in hand, hence, we resolve to imitation learning techniques due to ease of manually performing such tasks like manipulation that could be comfortably mimicked by our agent. Deploying current imitation learning algorithms on soft robotic systems have been observed to provide satisfactory results but there are still challenges in doing so. This review article thus posits an overview of various such algorithms along with instances of them being applied to real world scenarios and yielding state-of-the-art results followed by brief descriptions on various pristine branches of DRL research that may be centers of future research in this field of interest.

Modelling both long and short-term user interests from historical data is crucial for accurate recommendations. However, unifying these metrics across multiple application domains can be challenging, and existing approaches often rely on complex, intertwined models which can be difficult to interpret. To address this issue, we propose a lightweight, plug-and-play interest enhancement module that fuses interest vectors from two independent models. After analyzing the dataset, we identify deviations in the recommendation performance of long and short-term interest models. To compensate for these differences, we use feature enhancement and loss correction during training. In the fusion process, we explicitly split long-term interest features with longer duration into multiple local features. We then use a shared attention mechanism to fuse multiple local features with short-term interest features to obtain interaction features. To correct for bias between models, we introduce a comparison learning task that monitors the similarity between local features, short-term features, and interaction features. This adaptively reduces the distance between similar features. Our proposed module combines and compares multiple independent long-term and short-term interest models on multiple domain datasets. As a result, it not only accelerates the convergence of the models but also achieves outstanding performance in challenging recommendation scenarios.

The use of Information Communication Technologies (ICT) in education brings up new possibilities of promoting the learning and health experiences. In this sense, education contexts of 21st century must consider these two areas of knowledge, especially their integration. This article presents learning practices developed with mobile devices and games, in order to improve learning and well-being in children and adolescents undergoing cancer treatment in non-formal educational setting. The methodology is based on qualitative case studies with content-based data analysis, involving informal interviews and observation methods. The study considers data from 5 patients who participated in the research between 2015 and 2019. The results demonstrate a positive influence of the practices with mobile technologies and games in terms of learning and in the well-being feeling of patients during the treatment.

The application of machine learning techniques to satellite imagery has been the subject of interest in recent years. The increase in quality and quantity of images, made available by Earth observation programs, such as the Copernicus program, led to the generation of large amounts of data. Among the various applications of this data is the creation of land cover maps. The present work aimed to create machine learning models capable of accurately segment and classify satellite images, to automatically generate a land cover map of the Portuguese territory. Several experiments were carried out with the spectral bands of the Sentinel-2 satellite, with vegetation indices, and with several sets of land cover classes. Three machine learning architectures were evaluated, which adopt two different techniques for image classification. One of the classification techniques follows an object-oriented approach, and in this case the architecture adopted in our models was a U-Net artificial neural network. The other classification technique is pixel-oriented, and the machine learning models tested were random forest and support vector machine. The overall accuracy of the results obtained ranged from 68.6% to 94.75%, depending strongly on the number of classes into which the land cover is classified. The result of 94.75% was obtained when classifying the land cover only into 5 classes. However, a very interesting accuracy of 92.37% was achieved by the model when trained to classify 8 classes. These results are superior to those reported in the related bibliography.

Over a billion people live in slums, with poor sanitation, education, property rights and working conditions having direct impact on current residents and future generations. A key problem in relation to slums is slum mapping. Without delineations of where all slum settlements are, informed decisions cannot be made by policymakers in order to benefit the most in need. Satellite images have been used in combination with machine learning models to try and fill the gap in data availability of slum locations. Deep learning has been used on RGB images with some success but since labeled satellite images of slums are relatively low quality and the physical/visual manifestation of slums significantly varies within and across countries, it is important to quantify the uncertainty of predictions for reliable application in downstream tasks. Our solution is to train Monte Carlo dropout U-Net models on multispectral 13-band Sentinel-2 images from which we can calculate pixelwise epistemic (model) and aleatoric (data) uncertainty in our predictions. We trained our model on labelled images of Mumbai and verified our epistemic and aleatoric uncertainty quantification approach using altered models trained on modified datasets. We also used SHAP values to investigate how the different features contribute towards the model’s predictions and this showed that certain short-wave infrared and red-edge image bands are powerful features for determining the locations of slums within images. Having created our model with uncertainty quantification, in the future it can be applied to downstream tasks and decision-makers will know where predictions have been made with low uncertainty, giving them greater confidence in its deployment.

Ransomware attacks have rapidly proliferated, inflicting severe financial damages on businesses and individuals. Machine learning approaches to automate ransomware detection have shown promise but grapple with challenges like limited training data. This study introduces a novel deep learning model for few-shot ransomware classification. The model employs entropy features derived directly from malware binaries coupled with a twin neural network architecture utilizing transfer learning. Tests on over 1000 samples across 11 families demonstrate a weighted F1-score of 85.8\%, surpassing existing methods. The approach mitigates biases in limited training data and preserves intricacies lost in image-based features. It exhibits precise classification capabilities even with sparse samples of new ransomware variants. The research highlights the potential of entropy-driven deep learning to equip defenses against emerging zero-day ransomware strains.

Automatic ship detection provides an essential function towards maritime domain awareness for security or economic monitoring purposes. This work presents an approach for training a deep learning ship detector in Sentinel-2 multispectral images with few labeled examples. We design a network architecture for detecting ships with a backbone that can be pre-trained separately. By using Self Supervised Learning, an emerging unsupervised training procedure, we learn good features on Sentinel-2 images, without requiring labeling, to initialize our network’s backbone. The full network is then fine-tuned to learn to detect ships in challenging settings. We evaluated this approach versus pre-training on ImageNet and versus a classical image processing pipeline. We examined the impact of variations in the self-supervised learning step and we show that in the few-shot learning setting self-supervised pre-training achieves better results than ImageNet pre-training. When enough training data is available, our self-supervised approach is as good as ImageNet pre-training. We conclude that a better design of the self-supervised task and bigger non-annotated dataset sizes can lead to surpassing ImageNet pre-training performance without any annotation costs.

Plasma technology shows tremendous potential for revolutionizing oncology research and treatment. Reactive oxygen and nitrogen species, electromagnetic emissions generated through gas plasma jets, have attracted significant attention due to their selective cytotoxicity towards cancer cells. To leverage the full potential of plasma medicine, researchers have explored the use of mathematical models and various subsets of machine learning, such as reinforcement learning, and deep learning. This review emphasizes the significant application of AI algorithms in the adaptive plasma system, paving the way for precision and dynamic cancer treatment. Realizing the full potential of AI in plasma medicine, requires research efforts, data sharing and interdisciplinary collaborations. Unravelling the complex mechanisms, developing real-time diagnostics, and optimizing AI models will be crucial to harness the true power of plasma technology in oncology. The integration of personalized and dynamic plasma therapies, alongside AI and diagnostic sensors, presents a transformative approach to cancer treatment with the potential to improve outcomes globally.

State-of-the-art remote sensing scene classification methods employ different Convolutional Neural Network architectures for achieving very high classification performance. A trait shared by the majority of these methods is that the class associated with each example is ascertained by examining the activations of the last fully connected layer, and the networks are trained to minimize the cross-entropy between predictions extracted from this layer and ground-truth annotations. In this work, we extend this paradigm by introducing an additional output branch which maps the inputs to low dimensional representations, effectively extracting additional feature representations of the inputs. The proposed model imposes additional distance constrains on these representations with respect to identified class representatives, in addition to the traditional categorical cross-entropy between predictions and ground-truth. By extending the typical cross-entropy loss function with a distance learning function, our proposed approach achieves significant gains across a wide set of benchmark datasets in terms of classification, while providing additional evidence related to class membership and classification confidence.

This paper presents a study of an automated system for identifying planktic foraminifera at the species level. The system uses a combination of deep learning methods, specifically Convolutional Neural Networks (CNNs), to analyze digital images of foraminifera taken at different illumination angles. The dataset is composed of 1437 groups of sixteen grayscale images, one group for each foraminifer, that are then converted to RGB images with various processing methods. These RGB images are fed into a set of CNNs, organized in an Ensemble Learning (EL) environment. The ensemble is built by training different networks using different approaches for creating the RGB images. The study finds that an ensemble of CNN models trained on different RGB images improves the system's performance compared to other state-of-the-art approaches. The proposed system was also found to outperform human experts in classification accuracy.

The importance of radar-based human activity recognition increased significantly over the last two decades in safety and smart surveillance applications due to its superiority towards vision-based sensing in the presence of poor environmental conditions, like illumination, increased radiative heat, occlusion, and fog. An increased public sensitivity for privacy protection, and the progress of cost-effective manufacturing, led to a higher acceptance and distribution. Deep learning approaches proved that the manual feature extraction that relies heavily upon process knowledge can be avoided by its hierarchical, non-descriptive nature. On the other hand, ML techniques based on manual feature extraction provide a robust, yet empirical based approach where the computational effort is comparatively low. This review outlines the basics of classical ML- and DL-based human activity recognition and its advances while taking recent progress of both categories into regard. For every category, state-of-the-art methods are introduced, briefly explained and discussed in terms of their pros, cons and gaps. A comparative study is performed to evaluate the performance and computational effort based on a benchmarking data set to provide a common basis for the assessment of the techniques’ degree of suitability.

Accurate transfer learning of clinical outcomes, e.g., of the effects and side effects of drugs or other interventions, from one cellular context to another (in-vitro versus ex-vivo versus in-vivo, or across tissues), between cell-types, developmental stages, omics modalities or species, is considered tremendously useful. Ultimately, it may avoid most drug development failing in translation, despite large investments in the preclinical stages, which includes animal experiments requiring careful justification. Thus, when transferring a prediction task from a source (model) domain to a target domain, what counts is the high quality of the predictions in the target domain, requiring molecular states or processes common to both source and target that can be learned by the predictor, reflected by latent variables. These latent variables may form a compendium of knowledge that is learned in the source, to enable predictions in the target; usually, there are few, if any, labeled target training samples to learn from. Transductive learning then refers to the learning of the predictor in the source domain, transferring its outcome label calculations to the target domain, considering the same task. Inductive learning considers cases where the target predictor is performing a different yet related task as compared to the source predictor, making some labeled target data necessary. Often, there is also a need to first map the variables in the input/feature spaces (e.g. of gene names to orthologs) and/or the variables in the output/outcome spaces (e.g. by matching of labels). Transfer across omics modalities also requires that the molecular information flow connecting these modalities is sufficiently conserved. Only one of the methods for transfer learning we reviewed offers an assessment of input data, suggesting that transfer learning is unreliable in certain cases. Moreover, source domains feature their very own particularities, and transfer learning should consider these, e.g., as differences in pharmacokinetics, drug clearance or the microenvironment. In light of these general considerations, we here discuss and juxtapose various recent transfer learning approaches, specifically designed (or at least adaptable) to predict clinical (human in-vivo) outcomes based on molecular data, towards finding the right tool for a given task, and paving the way for a comprehensive and systematic comparison of the suitability and accuracy of transfer learning of clinical outcomes.

Text annotation is an essential element of the natural language processing approaches. The manual annotation process performed by humans has several drawbacks, such as subjectivity, slowness, fatigue, and possibly carelessness. In addition, annotators may annotate ambiguous data. So, we developed the concept of automated annotation to get the best annotations using several machine-learning approaches. The proposed approach is based on an ensemble algorithm of meta-learners and meta-vectorizer techniques. The approach employs a semi-supervised learning technique for automated annotation, aimed at detecting hate speech. This involves leveraging various machine learning algorithms, including Support Vector Machine (SVM), Decision Tree (DT), K-Nearest Neighbors (KNN), and Naive Bayes (NB), in conjunction with Word2Vec and TF-IDF text extraction methods. The annotation process is performed using 13,169 Indonesian YouTube comments data. The proposed model used a Stemming approach using data from Sastrawi and also new data of 2,245 words. Semi-supervised learning uses 5%, 10%, and 20% of labeled data as compared to performing labeling based on 80% of the datasets. In semi-supervised learning, the model learns from the labeled data, which provides explicit information, and the unlabeled data, which offers implicit insights. This hybrid approach enables the model to generalize and make informed predictions even when limited labeled data is available, ultimately enhancing its ability to handle real-world scenarios with scarce annotated information. In addition, the proposed method uses a variety of thresholds for matching words labeled with hate speech ranging from 0.6, 0.7, 0.8, and 0.9. The experiment showed that the KNN-Word2ec model has the best accuracy value of 96.9% with a scenario of 5%:80%:0.9. However, several other methods have also accuracy above 90%, such as SVM and DT based on both text extraction methods in several test scenarios.

In recent years, there has been a growing interest in the prediction of facial age due to its diverse applications in fields such as security, entertainment, and healthcare. The current scourge of underaged voting in Nigeria is a problem and this research delves into the realm of facial age prediction by employing four well-known convolutional neural network (CNN) architectures, namely VGG16, ResNet50, Mobile Net, and VGG19 to predict chronological ages from facial pictures of person. The objective is to achieve precise age estimation from facial images, utilizing two datasets: UTKFace and CASIA African Facial Datasets. The results of this investigation are noteworthy. Specifically, the VGG16 model which demonstrated remarkable performance, yielding a Mean Absolute Error (MAE) of 1.76 when applied to the UTK-Face dataset. Additionally, when utilizing Mobile-Net, an unprecedented MAE of 1.10 was achieved for the Casia Africa Face Dataset. Notably, this marks the first instance of employing the dataset for facial age detection with CNNs, and this approach outperformed previous works, yielding the lowest MAE among all the studies reviewed.

A key component for the proper functioning, availability, and reliability of power grids is the power trans- former. Although these are very reliable assets, the early detection of incipient degradation mechanisms are very important to prevent failures that may shorten their lifetime. In this work a review and comparative analysis, of classical Machine Learning algorithms (such as single and ensemble classification algorithms) and two automatic machine learning classifiers, are presented for the fault diagnosis of power transformers. The goal is to determine whether fully automated ML approaches are worse or better than traditional ML frameworks that require a human in the loop (such as a data scientist) to identify transformer faults from oil-dissolved gases data. The methodology uses a DB compiled from published and proprietary transformer fault samples. Faults data is obtained from the literature, the Duval pentagon method, and user-expert knowledge. The parameters from, either single or ensemble classifiers, were optimized through standard machine learning procedures. The results showed that the best performing algorithm is a robust automatic machine learning classifiers model followed by classical algorithms such as neural networks and stacking ensembles. These results highlight the ability of a robust automatic machine learning model for, handling imbalanced power transformers faults datasets with high accuracy, using the minimum tuning effort by the electrical experts. By providing the most probable transformer fault condition will reduce the time required to find and solve the fault.

Deep learning has taken over - both in problems beyond the realm of traditional, hand-crafted machine learning paradigms as well as in capturing the imagination of the practitioner sitting on top of petabytes of data. While the public perception about the efficacy of deep neural architectures in complex pattern recognition tasks grows, sequentially up-to-date primers on the current state of affairs must follow. In this review, we seek to present a refresher of the many different stacked, connectionist networks that make up the deep learning architectures followed by automatic architecture optimization protocols using multi-agent approaches. Further, since guaranteeing system uptime is fast becoming an indispensable asset across multiple industrial modalities, we include an investigative section on testing neural networks for fault detection and subsequent mitigation. This is followed by an exploratory survey of several application areas where deep learning has emerged as a game-changing technology - be it anomalous behavior detection in financial applications or financial time-series forecasting, predictive and prescriptive analytics, medical imaging, natural language processing or power systems research. The thrust of this review is on outlining emerging areas of application-oriented research within the deep learning community as well as to provide a handy reference to researchers seeking to embrace deep learning in their work for what it is: statistical pattern recognizers with unparalleled hierarchical structure learning capacity with the ability to scale with information.

Educational data-mining is an evolving discipline that focuses on the improvement of self-learning and adaptive methods. It is used for finding hidden patterns or intrinsic structures of educational data. In the arena of education, the heterogeneous data is involved and continuously growing in the paradigm of big-data. To extract meaningful information adaptively from big educational data, some specific data mining techniques are needed. This paper presents a clustering approach to partition students into different groups or clusters based on their learning behavior. Furthermore, personalized e-learning system architecture is also presented which detects and responds teaching contents according to the students’ learning capabilities. The primary objective includes the discovery of optimal settings, in which learners can improve their learning capabilities. Moreover, the administration can find essential hidden patterns to bring the effective reforms in the existing system. The clustering methods K-Means, K-Medoids, Density-based Spatial Clustering of Applications with Noise, Agglomerative Hierarchical Cluster Tree and Clustering by Fast Search and Finding of Density Peaks via Heat Diffusion (CFSFDP-HD) are analyzed using educational data mining. It is observed that more robust results can be achieved by the replacement of existing methods with CFSFDP-HD. The data mining techniques are equally effective to analyze the big data to make education systems vigorous.

In the present paper, an aerodynamic investigation of a high-speed train is performed. In the first section of this article, a generic high-speed train against a turbulent flow is simulated, numerically. The Reynolds-Averaged Navier-Stokes (RANS) equations combined with the SST turbulence model are applied to solve incompressible turbulent flow around a high-speed train. Flow structure, velocity and pressure contours and streamlines at some typical wind directions are the most important results of this simulation. The maximum and minimum values are specified and discussed. Also, the pressure coefficient for some critical points on the train surface is evaluated. In the following, the wind direction influence the aerodynamic key parameters as drag, lift, and side forces at the mentioned wind directions are analyzed and compared. Moreover, the effects of velocity changes (50, 60, 70, 80 and 90 m/s) are estimated and compared on the above flow and aerodynamic parameters. In the second section of the paper, various data-driven methods including Gene Expression Programming (GEP), Gaussian Process Regression (GPR), and random forest (RF), are applied for predicting output parameters. So, drag, lift and side forces and also minimum and a maximum of pressure coefficients for mentioned wind directions and velocity are predicted and compared using statistical parameters. Obtained results indicated that RF in all coefficients of wind direction and most coefficients of free stream velocity provided the most accurate predictions. As a conclusion, RF may be recommended for the prediction of aerodynamic coefficients.

Data Extraction is a technique is called as clustering which is used to retrieve data either from the files or data bases or both. This paper focuses on the performance evaluation parameters of the clustering algorithms based on different parameters or conditions or constraints and parameters which are used to perform the clustering process to get the clusters on the data sets. Therefore best clusters are retrieved when best parameters or conditions or constraints or preferences which are applied on the data sources for the clustering process. These parameters or conditions or constraints are opted by the user called as user preferences.

Deep learning dominates automated animal activity recognition (AAR) tasks due to the high performance on large-scale datasets. However, constructing centralised data across diverse farms raises data privacy issues. Federated learning (FL) provides a distributed learning solution to train a shared model by coordinating multiple farms (clients) without sharing their private data. Whereas directly applying FL to AAR tasks often faces two challenges: client-drift during local training and gradient conflicts during global aggregation. In this study, we develop a novel FL framework called FedAAR to achieve AAR with sensor data. Specifically, we devise a prototype-guided local update module to alleviate the client-drift issue, which introduces global prototypes as shared knowledge to force clients to learn consistent features. To reduce gradient conflicts between clients, we design a gradient refinement-based aggregation module to eliminate conflicting components between client gradients during global aggregation, thereby improving the agreement between clients. Experiments are conducted on a public dataset to verify FedAAR’s effectiveness, which consists of 87,621 2-s motion data. The results demonstrate that FedAAR outperforms state-of-the-arts, with precision (75.23%), recall (75.17%), F1-score (74.70%), and accuracy (88.88%), respectively. The ablation experiments also show FedAAR’s robustness against various factors (i.e., different data sizes, communication frequency, and client numbers).

The exploration and understanding of sentiments and emotions through textual analysis are foundational to advancements in natural language processing (NLP). While sentiment analysis, which involves discerning the polarity of text, has been extensively explored, the nuanced detection of complex emotions within textual content is still an emerging area of study. In this paper, we introduce EmoLeverage, a model leveraging Transformer architecture and enhanced with a novel Fusion Adapter module. This model aims to deepen emotion detection capabilities by drawing insights from fundamental sentiment analysis tasks, focusing exclusively on textual data. EmoLeverage sets a new standard in the field by demonstrating that deep, nuanced understanding of emotions is achievable through text alone, challenging and surpassing existing methodologies in emotion identification on key datasets. This is a significant step forward, considering the complexity of human emotions and the subtleties involved in their expression through language. By concentrating on textual input, EmoLeverage offers a promising avenue for applications where audio or visual data is not available, highlighting the versatility and potential of text-based emotional analysis. Our approach illustrates the untapped potential of leveraging sophisticated NLP techniques to enhance the detection and comprehension of a wide array of emotions from text. The success of EmoLeverage in outperforming conventional models in emotion recognition tasks underscores the model's effectiveness in parsing and understanding the intricate expressions of human emotions through linguistic cues alone. This achievement paves the way for more accurate, efficient, and nuanced emotion detection tools, potentially transforming various sectors by providing deeper insights into human emotional states through text.

We propose a novel transfer learning method for speech emotion recognition allowing us to obtain promising results when only few training data is available. With as low as 125 examples per emotion class, we were able to reach a higher accuracy than a strong baseline trained on 8 times more data. Our method leverages knowledge contained in pre-trained speech representations extracted from models trained on a more general self-supervised task which doesn’t require human annotations, such as the wav2vec model. We provide detailed insights on the benefits of our approach by varying the training data size, which can help labeling teams to work more efficiently. We compare performance with other popular methods on the IEMOCAP dataset, a well-benchmarked dataset among the Speech Emotion Recognition (SER) research community. Furthermore, we demonstrate that results can be greatly improved by combining acoustic and linguistic knowledge from transfer learning. We align acoustic pre-trained representations with semantic representations from the BERT model through an attention-based recurrent neural network. Performance improves significantly when combining both modalities and scales with the amount of data. When trained on the full IEMOCAP dataset, we reach a new state-of-the-art of 73.9% unweighted accuracy (UA).

Accurate and efficient medicinal plant image classification is of utmost importance as these plants produce a wide variety of bioactive compounds that offer therapeutic benefits. With a long history of medicinal plant usage, different parts of plants, such as flowers, leaves, and roots, have been recognized for their medicinal properties and are used for plant identification. However, leaf images are extensively used due to their convenient accessibility and are a major source of information. In recent years, transfer learning and fine-tuning, which use pre-trained deep convolutional networks to extract pertinent features, has emerged as an extremely effective approach for image identification problems. This study is leveraging the power of three component deep convolutional neural networks, namely VGG16, VGG19 and DenseNet201, to derive features from the input images of the medicinal plant dataset, containing leaf images of 30 classes. The models were compared and ensembled to make four hybrid models to enhance the predictive performance by utilizing the averaging and weighted averaging strategies. Quantitative experiments were carried out to evaluate the models on the Mendeley medicinal leaf dataset. The resultant ensemble of VGG19+DensNet201 with fine-tuning showcases enhanced capability in identifying medicinal plant images with an improvement of 7.43% and 5.8% compared with VGG19 and VGG16. Furthermore, VGG19+DensNet201 can outperform its standalone counterparts by achieving an accuracy of 99.12% on the test set. A thorough assessment with metrics such as accuracy, recall, precision, and f1-score firmly establishes the effectiveness of the ensemble strategy.

One of the greatest COVID-19 related challenges for children and their families was managing distance learning due to school closures. We also know from previous research that families with a child with a neurodevelopmental disorder such as ADHD or ASD were struggling more than others, but also experiences some positive effects. However, few qualitative studies have been conducted. The present study therefore aimed to investigate parental experiences of negative and positive effects of distance learning during the COVID-19 pandemic in a large sample of families with an adolescent with ADHD and/or ASD and a matched comparison group (n = 682). Data was collected through open-ended questions as part of a larger survey study. Five main themes, with different sub-themes were identified for both negative and positive effects: 1) Teaching, 2) Social, 3) Support, 4) Child factors, and 5) Home environment. In addiction, the main themes “Technical problems” were identified for negative effects. Families with a child with ADHD/ASD reported negative effects related to “Child factors” and “Support” more frequently than the controls, as well as negative effects related to some aspects of “Teaching” and “Technical problems”. Regarding positive effects, significant group differences were primarily found for the theme “Child factors”. These findings are discussed botn in terms of how to best prepare for possible future pandemics, but also how to best provide educational support for children with ADHD and/or ASD when schools are open.

Laboratory medicine has evolved from a mainly manual profession, providing few selected test results to a highly automated and standardized medical discipline, generating millions of test results per year. As next inevitable evolutional step, artificial intelligence (AI) algorithms will need to assist us in structuring and making sense of the masses of diagnostic data collected today. Such systems will be able to connect clinical and diagnostic data and to provide valuable suggestions in diagnosis, prognosis or therapeutic options. They will merge the often so separated worlds of the laboratory and the clinics. When used correctly, it will be a tool, capable of freeing the physicians time so that he/she can refocus on the patient. In this narrative review I therefore aim to provide an overview of what AI is, what applications currently are available in healthcare and in laboratory medicine in particular. I will discuss the challenges and pitfalls of applying AI algorithms and I will elaborate on the question if healthcare workers will be replaced by such systems in the near future.

The present paper explored the possibility of discriminating between different malignant degrees of Chondrosarcoma (CS), a diffused bone tumor, and Enchondroma (EC), its benign version, through an approach based on the coupling between Confocal Raman Microscopy (CRM) and Machine Learning (ML) techniques. Recently, Raman Spectroscopy has proven to be a powerful tool for a complete grading of CS by distinguishing between grade 1, 2 and 3, and to distinguish CS from EC. In this paper, we tested some models, belonging either to ML or to the sub-type called Deep Learning (DL), showing excellent classification performances, especially the DL algorithms, with classification accuracy approaching to 100%, making the models promising for future implementations of Raman spectroscopy when applied to oncology diagnosis. In turn, we highlighted how some proper ML models result in worse classification performances but better resolution of specific chemical target compounds, possible candidate to become malignant markers, with relevant implication for a correct diagnosis.

This article presents a review of current advances and future prospects in the field of forecasting renewable energy generation using machine learning (ML) and deep learning (DL) techniques. With the increasing penetration of renewable energy sources (RES) into the electricity grid, accurate forecasting of their generation becomes crucial for efficient grid operation and energy management. Traditional forecasting methods have limitations, and thus ML and DL algorithms have gained popularity due to their ability to learn complex relationships from data and provide accurate predictions. This paper reviews the different approaches and models that have been used for renewable energy forecasting and discusses their strengths and limitations. It also highlights the challenges and future research directions in the field, such as dealing with uncertainty and variability in renewable energy generation, data availability, and model interpretability. Finally, this paper emphasizes the importance of developing robust and accurate renewable energy forecasting models to enable the integration of RES into the electricity grid and facilitate the transition towards a sustainable energy future.

SLI and SLD constitute two autonomous neurodevelopmental disorders, which frequently cause challenges in the diagnosis process, especially due to their nature, which has caused disagreement among clinicians regarding their recognition as separate or overlapping disorders. The objective of the study was to enlighten the path of valid diagnosis and intervention during adolescence when the two disorders change their manifestation and overlap. 200 Greek adolescents (140 boys and 60 girls), 124 already diagnosed with SLD and 76 diagnosed with SLI, 12-16 years old, participated in the study. All participants were assessed in reading, oral and written language and mathematics (mathematical operations and mathematical reasoning) along with IQ testing. In order to determine statistically significant differences, the chi-square test, independent samples t-test, odd ratios and their 95 per cent confidence intervals were implemented. The results revealed that the SLI group presented significantly greater difficulties than SLD in their overall cognitive-mental profile and in most language and mathematical measurements (number concept, executive-procedural part of solving operations and mathematical reasoning). The similarity of the two groups was mainly detected in their deficient metacognitive, metalinguistic and metamnemonic strategies. The research concludes that SLD adolescents managed to overcome their difficulties to a significant degree, while adolescents with SLI still struggle with many learning areas.

Developing efficient Artificial Intelligence (AI)-enabled system to substitute human role in non-destructive testing is an emerging topic of considerable interest. In this study, we propose a novel impact-echo analysis system using online machine learning, which aims at achieving near-human performance for assessment of concrete structures. Current computerized impact-echo systems commonly employ lab-scale data to validate the models. In practice, however, the echo patterns can be far more complicated due to varying geometric shapes and materials of structures. To deal with a large variety of unseen data, we propose a sequential treatment for echo characterization. More specifically, the proposed system can adaptively update itself to approaching human performance in impact-echo data interpretation. To this end, a two-stage framework has been introduced, including echo feature extraction and the model updating scheme. Various state-of-the-art online learning algorithms have been reviewed and evaluated for the task. To conduct experimental validation, we collected 10,940 echo instances from multiple inspection sites with each sample had been annotated by human experts with healthy/defective condition labels. The results demonstrated that the proposed scheme achieved favorable echo pattern classification accuracy with high efficiency and low computation load.

Neural networks present the characteristics that the results are strongly dependent on the training data, the weight initialisation, and the hyper-parameters chosen. The determination of the distribution of a statistical estimator, as the Mean Squared Error (MSE) or the accuracy, is fundamental to evaluate the performance of a neural network model (NNM). For many machine learning models, as linear regression, it is possible to analytically obtain information as variance or confidence intervals on the results. Neural networks present the difficulty of being not analytically tractable due to their complexity. Therefore, it is impossible to easily estimate distributions of statistical estimators. When estimating the global performance of an NNM by estimating the MSE in a regression problem, for example, it is important to know the variance of the MSE. Bootstrap is one of the most important resampling techniques to estimate averages and variances, between other properties, of statistical estimators. In this tutorial, the application of two resampling (including bootstrap) techniques to the evaluation of neural networks’ performance is explained from both a theoretical and practical point of view. Pseudo-code of the algorithms is provided to facilitate their implementation. Computational aspects, as the training time, are discussed since resampling techniques always require to run simulations many thousands of times and, therefore, are computationally intensive. A specific version of the bootstrap algorithm is presented that allows the estimation of the distribution of a statistical estimator when dealing with an NNM in a computationally effective way. Finally, algorithms are compared on synthetically generated data to demonstrate their performance.

Log collection and storage is a crucial process for enterprises around the globe. Log analysis helps identify potential security breaches and, in some cases, is required by law for compliance. However, enterprises often delegate these responsibilities to third-party Cloud Service Providers (CSPs), where the logs are collected and processed for anomaly detection and stored in a data warehouse for archiving. Prevalent schemes rely on plain (unencrypted) data for anomaly detection. More often, these logs can reveal sensitive information about an organization or the customers of that organization. Hence, it is best to keep it encrypted at all times. This paper presents "SigML++," an extension of work done in "SigML." We utilize Fully Homomorphic Encryption (FHE) with the Cheon-Kim-Kim-Song (CKKS) scheme for supervised log anomaly detection on encrypted data. We use an Artificial Neural Network (ANN) based probabilistic polynomial approximations using a Perceptron with linear activation. We probabilistically approximate the Sigmoid activation function (σ(x)) in the encrypted domain for the intervals [−10,10] and [−50,50]. Experiments show better approximations for Logistic Regression (LR) and Support Vector Machine (SVM) for low-order polynomials.

This paper presents the development of an underwater crack detection system for structural integrity assessment of submerged structures, like offshore oil and gas installations, underwater pipelines, underwater foundations for bridges, dams etc. Focus is on use of machine learning based approaches. First a detailed literature review of state of the current methods for underwater surface crack detection is presented highlighting challenges and opportunities. An overview of image augmentation approach for creation of underwater optical effects is also presented. Experimental results using standard network based machine learning approach, used for surface crack detection in onshore environment, is presented. Series of Test cases are presented where existing networks performance are improved using augmented images for underwater conditions. The experimental results demonstrate the effectiveness and accuracy of the proposed system in detecting cracks in underwater structures. The system has the potential to improve the safety and reliability of underwater structures and prevent catastrophic failures.

In recent years, the viability of employing multi-agent reinforcement learning technology for adaptive traffic signal control has been extensively validated. However, owing to restricted communication among agents and the partial observability of the traffic environment, the process of mapping road network states to actions encounters numerous challenges. To address this problem, this paper proposes a multi-agent deep reinforcement learning model with an emphasis on communication content (CMARL). The model decouples the complex relationships between multi-signal agents through centralized training and decentralized execution. Specifically, we first pass the traffic state through an improved deep neural network to achieve the extraction of high-dimensional semantic information and the learning of the communication matrix. Then the agents selectively interact with each other based on the learned communication matrix and generate the final state features. Finally, the features are inputted to the QMIX network to achieve the final action selection. We compare the CMARL model with 6 other baseline algorithms in real traffic networks. The results show that CMARL can significantly reduce vehicle congestion, and run stably in various scenarios.

This work develops a new method for vector data augmentation. The proposed method applies the Principal Component Analysis (PCA) and determines the Eigenvectors of a set of training vectors for a Machine Learning (ML) method and uses them to generate the distilled vectors. The training and the PCA distilled vectors have the same dimension. The user chooses the number of vectors to be distilled and augmented to the set of training vectors. A statistical approach determines the lowest number of vectors to be distilled such that when augmented to the original vectors the extended set trains a ML classifier to achieve a required accuracy. Hence, the novelty of this study is, the distillation of vectors with the PCA method and their use to augment the original set of vectors. The advantage that comes from the novelty is that it increases the statistics of ML classifiers. To validate the advantage, we conducted experiments with four public databases and applied four classifiers: Neural Network, Logistic Regression, Support Vector Machine with linear and polynomial kernels. For the purpose of augmentation, we conduct several distillations including nested (double distillation). The latter notion means that new vectors were distilled from already distilled vectors. We train the classifiers with three sets of vectors: original vectors; original augmented with vectors distilled by PCA; original augmented with distilled and double distilled, by PCA, vectors. The experimental results are presented in the paper, and they confirm the advantage that the PCA distilled vectors increase the classification statistics of ML methods, if the distilled vectors augment the original training vectors.

This paper presents a case study analyzing which learning cognitive strategies are employed by students who play digital games in both in-school and out-of-school learning contexts. As a Case Study, it falls within the qualitative research category. The researcher integrated themselves to the subjects' learning facility for data-gathering, following all recommended ethical protocols for research involving underage subjects. The following data collection instruments were used: research diary, questionnaire, and interviews, conducted with teenagers who play digital games during formal learning activities. Analysis involved data-crossing and checking our data against existing studies and theories, as well as a cartographic production. Data indicates that digital games enhance student usage of cognitive strategies and favor competencies and abilities-based learning over formal learning contents.

In this paper, we redefine the boundaries of blood cell classification, expanding from 5 to 11 distinct blood cell types using the challenging 11-class PBC dataset. This shift enables a deeper understanding of blood cell diversity, overcoming previous limitations in medical image analysis. Our approach combines state-of-the-art deep learning techniques, including pre-trained ConvNets, ViTb16 models, and custom CNN architectures. We employ transfer learning, fine-tuning, and ensemble strategies, such as CBAM and Averaging ensembles, to achieve unprecedented accuracy and interpretability. Our fully fine-tuned EfficientNetV2 B0 model sets a new standard, with a macro-average precision, recall, and F1-score of 91%, 90%, and 90%, respectively, and an average accuracy of 93%. This breakthrough underscores the transformative potential of 11-class blood cell classification for more precise medical diagnoses. Moreover, our groundbreaking "Naturalize" augmentation technique produces remarkable results. The 2K-PBC dataset generated with "Naturalize" boasts a macro-average precision, recall, and F1-score of 97%, along with an average accuracy of 96% when leveraging the fully fine-tuned EfficientNetV2 B0 model. This innovation not only elevates classification performance but also addresses data scarcity and bias in medical deep learning. Our research marks a paradigm shift in blood cell classification, enabling more nuanced and insightful medical analyses. The "Naturalize" technique’s impact extends beyond blood cell classification, emphasizing the vital role of diverse and comprehensive datasets in advancing healthcare applications through deep learning.

Over the past decade, recognition of human activities (HAR) has become a vibrant field of research, in particular, the spread in our everyday lives of electronics such as mobile phones, smart cell phones, and video cameras. Furthermore, the advancement in the field of deep methodologies and other paradigms have enabled scientists to enable HAR in many areas, consisting of activities in fitness and wellness. For instance, HAR is one of many resorts to support older people through day-to-day activities to support their cognition and physicality. This study is centered on the key aspects deep learning plays in the development of HAR applications. Although numerous HAR examination studies were carried out previously, there have been no overall studies on this subject, in all the earlier studies there were only specific HAR-related subjects. A detailed review covering all the main subjects in this area is therefore essential. This study discusses the latest developments and works in HAR. It separates the methods and the advantages and disadvantages of each method group. This paper finally addresses many problems in the current HAR subject and provides recommendations for potential study.

This paper presents a study and implementation of a convolutional neural network to identify and recognize humpback whale specimens from the unique patterns of their tails. Starting from a dataset composed of images of whale tails, all the phases of the process of creation and training of a neural network are detailed – from the analysis and pre-processing of images to the elaboration of predictions, using TensorFlow and Keras frameworks. Other possible alternatives are also explained when it comes to tackling this problem and the complications that have arisen during the process of developing this paper.

Accurate prediction of crop production is essential in effectively managing agricultural countries' food security and economic resilience. This study evaluates the performance of statistical and machine learning-based methods for large-scale crop production forecasting. We predict the quarterly production of 325 crops (including fruits, vegetables, cereals, non-food, and industrial crops) across 83 provinces in the Philippines. Using a comprehensive dataset of 10,949 time series over 13 years, we demonstrate that a global forecasting approach using a state-of-the-art deep learning architecture, the transformer, significantly outperforms traditional local forecasting approaches built on statistical and baseline methods. By leveraging cross-series information, our proposed way is scalable and works well even with time series that are short, sparse, intermittent, or exhibit structural breaks/regime shifts. The results of this study further advance the field of applied forecasting in agricultural production and provide a practical and effective decision-support tool for policymakers that oversee the farm sector on a national scale.

This paper explores the potential of using the SAM segmentator to enhance the segmentation capability of known methods. SAM is a promptable segmentation system that offers zero-shot generalization to unfamiliar objects and images, eliminating the need for additional training. The open-source nature of SAM on GitHub allows for easy access and implementation. In our experiments, we aim to improve the segmentation performance by providing SAM with checkpoints extracted from the masks produced by DeepLabv3+, then merging the segmentation masks provided by these two networks. Additionally, we examine the \enquote{oracle} method (as upper bound baseline performance), where segmentation masks are inferred only by SAM with checkpoints extracted from ground truth. In addition, we tested in the CAMO datasets an ensemble of PVTv2 transformers; combining the ensemble and SAM yields state-of-the-art performance in that dataset. The results of our study provide valuable insights into the potential of incorporating the SAM segmentator into existing segmentation techniques. We release with this paper the open-source implementation of our method.

Electroencephalography (EEG) is a brain imaging technique in which electrodes are placed on the scalp. EEG signals are commonly decomposed into frequency bands called delta, theta, alpha, and beta.While these bands have been shown to be useful for characterizing various brain states, their utility as a one-size-fits-all analysis tool remains unclear. We present a two-part data-driven methodology for objectively determining the best EEG bands for a given dataset in this paper. First, a decision tree is used to estimate the optimal frequency band boundaries for reproducing the signal’s power spectrum for a predetermined number of bands. The optimal number of bands is then determined using an Akaike Information Criterion (AIC)-inspired quality score that balances goodness-of-fit with a small band count. Data-driven EEG band discovery may aid in objectively capturing key signal components and uncovering new indices of brain activity.

Heart failure is a prevalent cardiovascular condition with significant health implications, necessitating effective diagnostic strategies for timely intervention. This study explores the potential of continuous monitoring of non-invasive signals, specifically integrating Photoplethysmogram (PPG) and Electrocardiogram (ECG), for enhancing early detection and diagnosis of heart failure. Leveraging a dataset from the MIMIC-III database, encompassing 682 heart failure patients and 954 controls. Feature selection techniques were used to systematically select key features which were identified for their clinical relevance and significance in capturing cardiovascular dynamics and to reduce computational complexity and to decrease the chance of overfitting the ML algorithms. These features are then utilized to train and evaluate machine learning algorithms, resulting in a model with an impressive accuracy of 98%, sensitivity of 97.60%, specificity of 96.90%, and precision of 97.20%. The integrated approach outperforms single-signal strategies, showcas-ng its potential for early, precise, and non-invasive heart failure diagnosis. Furthermore, the study underscores the significance of continuous monitoring through wearables, emphasizing the benefits of integrating multiple signals for a comprehensive evaluation of cardiovascular health. The proposed approach holds promise for implementation in hardware systems to enable continuous monitoring, aiding in early detection and prevention of critical health conditions.

(1) Purpose: This work aims at developing a generalizable MRI reconstruction model in the meta-learning framework. The standard benchmarks in meta-learning are challenged by learning on diverse task distributions. The proposed network learns the regularization function in a variational model and reconstructs MR images with various under-sampling ratios or patterns that may or may not be seen in the training data by leveraging a heterogeneous dataset. (2) Methods: We propose an unrolling network induced by learnable optimization algorithms (LOA) for solving our nonconvex nonsmooth variational model for MRI reconstruction. In this model, the learnable regularization function contains a task-invariant common feature encoder and task-specific learner represented by a shallow network. To train the network we split the training data into two parts: training and validation, and introduce a bilevel optimization algorithm. The lower-level optimization trains task-invariant parameters for the feature encoder with fixed parameters of the task-specific learner on the training dataset, and the upper-level optimizes the parameters of the task-specific learner on the validation dataset. (3) Results: The PSNR increases 1.5 dB on average compared to the network trained through conventional supervised learning on the seen CS ratios. We test the result of quick adaption on the unseen tasks after meta-training, the average PSNR arises 1.22 dB compared to the conventional learning procedure that is directly trained on the unseen CS ratios in the meanwhile saving half of the training time. The average PSNR arises 1.87 dB for unseen sampling patterns comparing to conventional learning; (4) Conclusion: We proposed a meta-learning framework consisting of the base network architecture, design of regularization, and bi-level optimization-based training. The network inherits the convergence property of the LOA and interpretation of the variational model. The generalization ability is improved by the designated regularization and bilevel optimization-based training algorithm.

This paper presents a non-random weight initialization method in convolutional layers of neural networks examined with the Fast Gradient Sign Method (FSGM) attack. This paper's focus is convolutional layers, and are the layers that have been responsible for better than human performance in image categorization. The proposed method induces earlier learning through the use of striped forms, and as such has less unlearning of the existing random number speckled methods, consistent with the intuitions of Hubel and Wiesel. The proposed method provides a higher performing accuracy in a single epoch, with improvements of between 3-5% in a well known benchmark model, of which the first epoch is the most relevant as it is the epoch after initialization. The proposed method is also repeatable and deterministic, as a desirable quality for safety critical applications in image classification within sensors. That method is robust to Glorot/Xavier and He initialization limits as well. The proposed non-random initialization was examined under adversarial perturbation attack through the FGSM approach with transferred learning, as a technique to measure the affect in transferred learning with controlled distortions, and finds that the proposed method is less compromised to the original validation dataset, with higher distorted datasets.

This paper offers a comprehensive examination of the process involved in developing and automating supervised end-to-end machine learning workflows for forecasting and classification purposes. It offers a complete overview of the components (i.e. feature engineering, model selection, etc), principles (i.e. bias-variance decomposition, model complexity, overfitting, model sensitivity to feature assumptions and scaling, output interpretability, etc), models (i.e. neural networks, regression models, etc), methods (i.e. Cross-Validation, data augmentation, etc), metrics (i.e. Mean Squared Error, F1-score, etc) and tools that rule most supervised learning applications with numerical and categorical data, as well as their integration, automation, and deployment. The end goal and contribution of this paper is the education and guidance of the non-AI expert academic community over complete and rigorous machine learning pipelines and data science practices, from problem scoping to design and state-of-the-art automation tools, including basic principles and reasoning in the choice of methods. The paper delves into the critical stages of supervised machine learning workflow development, many of which are often omitted by researchers due to brevity, and covers foundational concepts essential for understanding and optimizing a functional machine learning workflow, thereby offering a holistic view of task-specific application development for applied researchers who are not AI experts.

The accurate prediction of molecular properties such as lipophilicity and aqueous solubility are of great importance and pose challenges in several stages of the drug discovery pipeline. Machine learning methods like graph-based neural networks (GNNs) have shown exceptionally good performance in predicting these properties. In this work, we introduce a novel GNN architecture, called directed edge graph isomorphism network (D-GIN). It is composed of two distinct sub-architectures (D-MPNN, GIN) and achieves an improvement in accuracy over its sub-architectures employing various learning, and featurization strategies. We argue that combining models with different key aspects help make graph neural networks deeper and simultaneously increase their predictive power. Furthermore, we address current limitations in assessment of deep-learning models, namely, comparison of single training run performance metrics, and offer a more robust solution.

Alzheimer's is a disease for which there is no cure. Diagnosing Alzheimer's Disease (AD) early facilitates family planning and cost control. The purpose of this study is to predict the presence of AD using socio-demographic, clinical, and Magnetic Resonance Imaging (MRI) data. Early detection of AD enables family planning and may reduce costs by delaying long-term care. Accurate, non-imagery methods also reduce patient costs. The Open Access Series of Imaging Studies (OASIS-1) cross-sectional MRI data were analyzed. A gradient boosted machine (GBM) predicted the presence of AD as a function of gender, age, education, socioeconomic status (SES), and Mini-Mental State Exam (MMSE). A Residual Network with 50 layers (ResNet-50) predicted CDR presence and severity from MRI's (multi-class classification). The GBM achieved a mean 91.3% prediction accuracy (10-fold stratified cross validation) for dichotomous CDR using socio-demographic and MMSE variables. MMSE was the most important feature. ResNet-50 using image generation techniques based on an 80% training set resulted in 98.99% three class prediction accuracy on 4,139 images (20% validation set) at Epoch 133 and nearly perfect multi-class predication accuracy on the training set (99.34%). Machine Learning methods classify AD with high accuracy. GBM models may help provide initial detection based on non-imagery analysis, while ResNet-50 network models might help identify AD patients automatically prior to provider review.

Bayesian models allow us to investigate children’s belief revision alongside physiological states like “surprise”. Recent work finds that pupil dilation (or the “pupillary surprise response”) following expectancy-violations may be predictive of belief revision. How can probabilistic models inform interpretations of “surprise”? Shannon Information considers the likelihood of an observed event, given prior beliefs – suggesting stronger surprise occurs following unlikely events. In contrast, Kullback-Leibler divergence considers the “dissimilarity” between prior beliefs and updated beliefs following observations – with greater surprise indicating more change between belief states to accommodate information. To assess these accounts under different learning contexts, we use Bayesian models that compare these computational measures of “surprise” to contexts where children are asked to either predict or to evaluate the same evidence during a water displacement task. We find correlations between the computed Kullback-Leibler divergence and children’s pupillometry responses only when children actively make predictions, and no correlation between Shannon Information and pupillometry. This suggests that when children attend to their beliefs and make predictions, pupillary responses may signal the degree of divergence between a child’s current beliefs and updated, more accommodating beliefs.

This study examined the effect of an educational hybrid physical education (PE) intervention on cognitive performance and academic achievement in adolescents. A 9-month group-randomized controlled trial was conducted in 150 participants (age: 14.63 ± 1.38) allocated into control group (CG, n = 37) and experimental group (EG, n = 113). Inhibition, verbal fluency, planning and academic achievement were assessed. Significant differences were observed in the post-test for cognitive inhibition, verbal fluency in animals, and the average from verbal fluency in favour of the EG. With regard to the intervention, verbal fluency in animals, verbal fluency in vegetables, the average of verbal fluency, cognitive inhibition, language, the average of all subjects, the average of all subjects except PE, and the average from the core subjects) increased significantly in the EG. The last five variables (the academic ones and cognitive inhibition) also increased in the CG, in addition to mathematics. This study contributes to the knowledge by suggesting that both methodologies produced improvements in the measured variables, but the use of a hybrid program based on TPSR and gamification strategies produce improvements in cognitive performance, specifically through the cognitive inhibition and verbal fluency.

Gamification is the use of game mechanics in non-game situations. It is a solution to which businesses are turning more and more, since, among other things, it also responds to the individual's need for fun. What are the prevailing trends and to what extent can it be exploited in the period we are going through? This paper tries to contribute to a better understanding of gamification with main aim to help employees in their work, and businesses to achieve goals and be more competitive.

The latest methods based on deep learning have achieved amazing results regarding the complex work of inpainting large missing areas in an image. This type of method generally attempts to generate one single "optimal" inpainting result, ignoring many other plausible results. However, considering the uncertainty of the inpainting task, one sole result can hardly be regarded as a desired regeneration of the missing area. In view of this weakness, which is related to the design of the previous algorithms, we propose a novel deep generative model equipped with a brand new style extractor which can extract the style noise (a latent vector) from the ground truth image. Once obtained, the extracted style noise and the ground truth image are both input into the generator. We also craft a consistency loss that guides the generated image to approximate the ground truth. Meanwhile, the same extractor captures the style noise from the generated image, which is forced to approach the input noise according to the consistency loss. After iterations, our generator is able to learn the styles corresponding to multiple sets of noise. The proposed model can generate a (sufficiently large) number of inpainting results consistent with the context semantics of the image. Moreover, we check the effectiveness of our model on three databases, i.e., CelebA, Agricultural Disease, and MauFlex. Compared to state-of-the-art inpainting methods, this model is able to offer desirable inpainting results with both a better quality and higher diversity. The code and model will be made available on https://github.com/vivitsai/SEGAN.

With the popularity of solar energy in the electricity market, demand arises for data such as precise locations of solar panels for efficient energy planning, management, and distribution. However, this data is not easily accessible and in some cases, information such as precise locations does not exist. Furthermore, existing data sets for training semantic segmentation models of PV installations are limited, and their annotation is time-consuming and labor-intensive. Therefore, for additional remote sensing (RS) data creation, the pix2pix generative adversarial network (GAN) is utilized, enriching the original resampled training data of varying GSDs without compromising its integrity. Experiments done with the DeepLabV3 model, ResNet-50 backbone, and pix2pix GAN architecture were conducted to find the optimal configuration for an accurate RS imagery segmentation model. The result is a fine-tuned solar panel semantic segmentation model, trained using transfer learning and utilizing an optimal amount – 60% of generated RS imagery for additional training data, increasing model accuracy. The findings demonstrate the benefits of using GAN-generated images as additional training data, increasing the size of small data sets, and improving the capabilities of the segmentation model for solar panel detection in RS images.

Hepatocellular carcinoma is the most common primary malignant hepatic tumor and occurs most often in the setting of chronic liver disease. Liver transplantation is a curative treatment option and is an ideal solution because it solves the chronic underlying liver disorder while removing the malignant lesion. Because of organ shortages this treatment can only be applied to carefully selected patients according to clinical guidelines to minimize risk of recurrence. Artificial intelligence is an emerging technology with multiple applications in medicine with a predilection for domains that work with medical imaging like radiology. With the help of these technologies laborious tasks like segmentation can be automated and workflow in radiology departments can be improved. Other roles include in depth pixel-wise analysis of lesions by radiomics or deep learning in order to find new imaging criteria that allow better prediction of treatment response or risk of recurrence. Liver transplant is an ideal treatment for patients with hepatocellular carcinoma in the setting of chronic liver disease and artificial intelligence could provide solutions for improving the management of liver transplant candidates to improve survival.

Background and motivation: Combining Deep Reinforcement Learning (Deep RL) and Health Systems Simulations has significant potential, for both research into improving Deep RL performance and safety, and in operational practice. While individual toolkits exist for Deep RL and Health Systems Simulations, no framework to integrate the two has been established. Aim: Provide a framework for integrating Deep RL Networks with Health System Simulations, and to ensure this framework is compatible with Deep RL agents that have been developed and tested using OpenAI Gym. Methods: We developed our framework based on the OpenAI Gym framework, and demonstrate its use on a simple hospital bed capacity model. We built the Deep RL agents using PyTorch, and the Hospital Simulation using SimPy. Results: We demonstrate example models using a Double Deep Q Network or a Duelling Double Deep Q Network as the Deep RL agent. Conclusion: SimPy may be used to create Health System Simulations that are compatible with agents developed and tested on OpenAI Gym environments. GitHub repository of code: https://github.com/MichaelAllen1966/learninghospital

The extent of single and multi-cropping systems in any region, and potential changes to it, have consequences on food and resource use raising important policy questions. However, addressing these questions is limited by a lack of reliable data on multi-cropping practices at a high spatial resolution, especially in areas with high crop diversity. In this paper, we describe a relatively low-cost and scalable method to identify double cropping at the field-scale using satellite (Landsat) imagery. The process combines machine learning methods with expert labeling. We demonstrate the process by measuring double cropping extent in a portion of Washington State in the Pacific Northwest United States--- a region with significant production of more than 60 distinct types of crops including hay, fruits, vegetables, and grains in irrigated settings. Our results indicate that the current state-of-the-art methods for identifying cropping intensity---that apply rule-based thresholds on vegetation indices---do not work well in regions with high-crop-diversity. Our deep learning model was able to capture the diverse nuances and achieve a high accuracy (99\% overall accuracy and 0.92 Kappa coefficient). Our expert labeling process worked well and has potential as a relatively low-cost, scalable approach for remote sensing applications. The product developed here is valuable to inform several policy questions related to food production and resource use.

Artificial neural networks are a powerful tool for managing data that is difficult to process and interpret. This paper presents the study of artificial neural networks for information processing generated by an optical encoder based on the polarization of light. A machine learning technique is proposed to train the neural networks, such that the system can predict with remarkable accuracy the angular position in which the rotating element of the neuro-encoder is located, based on information provided by light’s phase shifting arrangements. The proposed neural designs show excellent performance in small angular intervals, and a methodology is proposed to avoid losing this remarkable characteristic in measurements from 0 to 180o or even to 360o. The neuro-encoder is implemented in simulation stage to obtain performance results. This study can be useful to improve capabilities of resolvers or other polyphasic sensors.

Abstract: Image aesthetic assessment (IAA) with neural attention has made significant progress due to its effectiveness in object recognition. Current studies have shown that the features learned by convolutional neural networks (CNN) at different learning stages indicate meaningful information. The shallow feature contains the low-level information of images and the deep feature perceives the image semantics and themes. Inspired by this, we propose a visual enhancement network with feature fusion (FF-VEN). It consists of two sub-modules, the visual enhancement module (VE module) and the shallow and deep feature fusion module (SDFF module). The former uses an adaptive filter in the spatial domain to simulate human eyes according to the region of interest (ROI) extracted by neural feedback. The latter not only takes out the shallow feature and the deep feature by transverse connection, but also uses a feature fusion unit (FFU) to fuse the pooled features together with the aim of information contribution maximization. Experiments on standard AVA dataset and Photo.net dataset show the effectiveness of FF-VEN.

The emergence of unmanned aerial vehicles (UAVs) raised multiple concerns, given their potential for malicious misuse in unlawful acts Vision-based counter-UAV applications offer a reliable solution compared to acoustic and radio frequency-based solutions because of their high detection accuracy in diverse weather conditions. The existing solutions work well on trained datasets, but their accuracy is relatively low for real-time detection. In this paper, we model deep learning-empowered solutions to improve the multi-class UAV's classification performance using single-shot object detection algorithms YOLOv5 and YOLOv7. The transfer learning is employed for performance improvement and rapid training with improved results. We customized a multi-class dataset containing multi-rotor, fixed-wing, and single-rotor UAVs in challenging weather conditions. Experiments show that the integration of transfer learning has achieved good results, with an overall best average-classification precision of 94\%, an average recall of 93.1\%, a mAP$@$0.5 average of 95.3\%, and an average F1 score of 92.33\%. The dataset and code are available as an open source: https://github.com/ZeeshanKaleem/YOLOV5-Large-vs-YOLOV7.git

Smartphones are prone to SMS phishing due to the rapid growth in the availability of smart mobile technologies driven by Internet connections. Also, detecting phishing SMS is a challenging task due to the unstructured nature of SMS text data with non-linear complex correlations. In this concern, considering the recent advancements in the domain of cybersecurity, we have proposed a hybrid deep learning framework that extracts robust features from SMS texts followed by an automatic detection of Phishing SMS. Due to combining the potential capability of individual models into one hybrid framework, it has outperformed various other individual machine learning and deep learning models. The proposed Phishing Detection framework is an effective hybrid combination of pretrained transformer model, MPNet (Masked and Permuted Language Modeling), with supervised ConvNets (CNN) and Bi-directional Gated Recurrent Units (GRU). It is intended to successfully detect unstructured short phishing text messages that contain complex patterns.

As artificial intelligence continues to advance, researchers are increasingly using machine learning algorithms to study the factors that make people more susceptible to phishing scams. Most studies in this area have taken one of two approaches: either they explore statistical associations between various factors and susceptibility, or they use complex models such as deep neural networks to predict phishing behavior. However, these approaches have limitations in terms of providing practical insights for individuals to avoid future phishing attacks and delivering personalized explanations regarding their susceptibility to phishing. In this paper, we propose a machine learning approach that leverages explainable artificial intelligence techniques to examine the influence of human and demographic factors on susceptibility to phishing attacks. Our analysis reveals that psychological factors such as impulsivity and conscientiousness, as well as appropriate online security habits, significantly affect an individual's susceptibility to phishing attacks. Furthermore, our individualized case-by-case approach offers personalized recommendations on mitigating the risk of falling prey to phishing exploits, considering the specific circumstances of each individual.

The work presented in this paper makes multiple scientific contributions related to the investigation of the global fear associated with COVID-19 by performing a comprehensive analysis of a dataset comprising survey responses of participants from 40 countries. First, the results of subjectivity analysis of responses where participants indicated their biggest concern related to COVID-19 showed that the average subjectivity in responses by the age group of 41-50 decreased from April 2020 to June 2020, the average subjectivity in responses by the age group of 71-80 drastically increased from May 2020, and the age group of 11-20 indicated the least level of subjectivity in their responses between June 2020 to August 2020. Second, subjectivity analysis also revealed the percentage of highly opinionated, neutral opinionated, and least opinionated responses per age-group where the analyzed age groups were 11-20, 21-30, 31-40, 41-50, 51-60, 61-70, 71-80, and 81-90. For instance, the percentage of highly opinionated, neutral opinionated, and least opinionated responses by the age group of 11-20 were 17.92%, 16.24%, and 65.84%, respectively. Third, data analysis of responses from different age groups showed that the highest percentage of responses indicating that they were very worried about COVID-19 came from individuals in the age group of 21-30. Fourth, data analysis of the survey responses also revealed that in the context of taking precautions to prevent contracting COVID-19, the percentage of individuals in the age group of 31-40 taking precautions was higher as compared to the percentages of individuals from the age groups of 41-50, 51-60, 61-70, 71-80, and 81-90. Finally, a deep learning model was developed to detect if the survey respondents were seeing or planning to see a psychologist or psychiatrist for any mental health issues related to COVID-19. The deep learning model used the responses to multiple questions in the context of fear, preparedness, and response related to COVID-19 from the dataset and achieved an overall performance accuracy of 91.62% after 500 epochs.

The explosion of data has transformed the world since much more information is available for collection and analysis than ever before. To extract valuable information from the data in different dimensions, various deep learning models have been developed in the past years. Although these models have demonstrated their strong capability on improving products and services in various applications, training them is still a time-consuming and resource-intensive process. Presently, cloud, one of the most powerful computing infrastructures, has been used for the training. However, how to manage cloud computing resources and to perform the training efficiently is still challenging current techniques. For example, general resource scheduling approaches, such as spread priority and balanced resource schedulers, actually do not work well with deep learning workloads. Besides, the resource allocation problem on a cluster can be divide into two subproblems: (1) local resource optimization: improve resource configuration for a single machine; (2) global resource optimization: improve the cluster-wide resource allocation. In this thesis, we propose two novel container schedulers, FlowCon and SpeCon, that are designed to address these two subproblems respectively and specifically to optimize performance of short-lived deep learning applications in the cloud. FlowCon focuses on resource configuration of single-node in a cluster, as show that it efficiently improves deep learning tasks completion time and resource utilization, and reduces the completion time of a specific job by up to 42.06\% without sacrificing the overall system time. SpeCon targets on cluster-wide resource configuration that speculatively migrate slow-growing models to release resources for fast-growing ones. Based on our experiments, SpeCon improves makespan for up to 24.7\%, compared to current approaches.

In the past years, various intelligent machine learning and deep learning algorithms have been developed and widely applied for gearbox fault detection and diagnosis. However, the real-time application of these intelligent algorithms has been limited, mainly due to the fact that the model developed using data from one machine or one operating condition has serious diagnosis performance degradation when applied to another machine or the same machine with a different operating condition. The reason for poor model generalization is the distribution discrepancy between the training and testing data. This paper proposes to address this issue using a deep learning based cross domain adaptation approach for gearbox fault diagnosis. Labelled data from training dataset and unlabeled data from testing dataset is used to achieve the cross-domain adaptation task. A deep convolutional neural network (CNN) is used as the main architecture. Maximum mean discrepancy is used as a measure to minimize the distribution distance between the labelled training data and unlabeled testing data. The study proposes to reduce the discrepancy between the two domains in multiple layers of the designed CNN to adapt the learned representations from the training data to be applied in the testing data. The proposed approach is evaluated using experimental data from a gearbox under significant speed variation and multiple health conditions. An appropriate benchmarking with both traditional machine learning methods and other domain adaptation methods demonstrates the superiority of the proposed method.

Free Floating Car Sharing (FFCS) services are a flexible alternative to car ownership. These transportation services show highly dynamic usage both over different hours of the day, and across different city areas. In this work, we study the problem of predicting FFCS demand patterns -- a problem of great importance to an adequate provisioning of the service. We tackle both the prediction of the demand i) over time and ii) over space. We rely on months of real FFCS rides in Vancouver, which constitute our ground truth. We enrich this data with detailed socio-demographic information obtained from large open-data repositories to predict usage patterns. Our aim is to offer a thorough comparison of several machine learning algorithms in terms of accuracy and easiness of training, and to assess the effectiveness of current state-of-art approaches to address the prediction problem. Our results show that it is possible to predict the future usage with relative errors down to 10%, and the spatial prediction can be estimated with relative errors of about 40%. Our study also uncovered the socio-demographic features that most strongly correlate with FFCS usage, providing interesting insights for providers opening service in new regions.

Medical health systems have been concentrating on artificial intelligence techniques for speedy diagnosis. However, the recording of health data in a standard form still requires attention so that machine learning can be more accurate and reliable by considering multiple parameters. The aim of this study is to develop a general framework for recording diagnostic data in an international standard format to facilitate auto-prediction of eye diseases. Efforts were made to ensure error-free data entry by developing a user-friendly interface. Furthermore, different machine learning algorithms were used to analyze patient data based on multiple parameters, including age, illness history, and clinical observations. This data was formatted according to structured hierarchies designed by medical experts, whereas diagnosis was made as per the ICD-10 coding developed by the American Academy of Ophthalmology. Furthermore, the system is designed to evolve through self-learning by adding new classifications for both diagnosis and symptoms. The classification results from tree-based methods demonstrated that the proposed framework performs satisfactorily given a sufficient amount of data. The random forest and decision tree algorithms predicted more accurately as compared to neural networks and the na{\"i}ve Bayes algorithm owing to structured data arrangement.

Deep learning heavily relies on statistical correlations to drive artificial intelligence (AI) innovations, particularly in computer vision applications like autonomous driving and robotics. However, despite providing a solid foundation for deep learning, these statistical correlations can be vulnerable to unforeseen and uncontrolled factors. The lack of prior knowledge guidance can result in spurious correlations, introducing confounding factors and affecting the model's robustness. To address this challenge, recent research efforts have focused on integrating causal theory into deep learning methodologies. By modelling the inherent and unbiased causal structure, causal theory can potentially mitigate the impact of spurious correlations effectively. Hence, this paper explores the basics of causal methodologies in image classification.

Several methods have been tried to estimate air temperature using satellite imagery. In this paper, the results of two machine learning algorithms, Support Vector Machine and Random Forest, are compared with Multivariate Linear Regression, TVX and Ordinary kriging. Several geographic, remote sensing and time variables are used as predictors. The validation is carried out using four different statistics on a daily basis allowing the use of ANOVA to compare the results. The main conclusion is that Random Forest with residual kriging produces the best results (R$^2$=0.612 $\pm$ 0.019, NSE=0.578 $\pm$ 0.025, RMSE=1.068 $\pm$ 0.027, PBIAS=-0.172 $\pm$ 0.046), whereas TVX produces the least accurate results. The environmental conditions in the study area are not really suited to TVX, moreover this method only takes into account satellite data. On the other hand, regression methods (Support Vector Machine, Random Forest and Multivariate Linear Regression) use several parameters that are easily calculated from a Digital Elevation Model, adding very little difficulty to the use of satellite data alone. The most important variables in the Random Forest Model were satellite temperature, potential irradiation and cdayt, a cosine transformation of the julian day.

The central goal of few-shot scene classification is to learn a model that can generalize well to a novel scene category (UNSEEN) from only one or a few labeled examples. Recent works in the remote sensing (RS) community tackle this challenge by developing algorithms in a meta-learning manner. However, most prior approaches have either focused on rapidly optimizing a meta-learner or aimed at finding good similarity metrics while overlooking the embedding power. Here we propose a novel Task-Adaptive Embedding Learning (TAEL) framework that complements the existing methods by giving full play to feature embedding’s dual roles in few-shot scene classification - representing images and constructing classifiers in the embedding space. First, we design a lightweight network that enriches the diversity and expressive capacity of embeddings by dynamically fusing information from multiple kernels. Second, we present a task-adaptive strategy that helps to generate more discriminative representations by transforming the universal embeddings into task-specific embeddings via a self-attention mechanism. We evaluate our model in the standard few-shot learning setting on two challenging datasets: NWPU-RESISC4 and RSD46-WHU. Experimental results demonstrate that, on all tasks, our method achieves state-of-the-art performance by a significant margin.

Squash is a sport where referee decisions are essential to the game. However, referee decisions in squash are very subjective in nature. Disputes, both from the players and the audience, regularly happen because the referee made a controversial call. In this study ,we proposed to automate the referee decision process through machine learning. We trained neural networks to predict squash referee decisions using data from 400 referee decisions acquired through extensive video footage reviewing and labeling. Six positional values were extracted, including the attacking player’s position, the retreating player’s position, the ball’s position in the frame, the ball’s projected first bounce, the ball’s projected second bounce, and the attacking player’s racket head position. We calculated nine additional distance values, such as the distance between players and the distance from the attacking player's racket head to the ball's path. Models were trained on Wolfram Mathematica and Python using these values. The best Wolfram Mathematica model achieved an 86% ± 3.03% accuracy, while the best Python model performed a 0.852 ± 0.051 accuracy (85.2% ± 5.1%). These accuracies surpass 85%, demonstrating near-human performances. Our model has great potential for improvement as it's currently trained with limited data (400 referee decisions) and lacks crucial data points such as time and speed. The performance of our model is almost surely going to improve significantly with a larger training data set. Unlike human referees, machine learning models follow a consistent standard, have unlimited attention spans, and make decisions instantly. If the accuracy is improved in the future, the model can potentially serve as an extra refereeing official for both professional and amateur squash matches. Both the analysis of referee decisions in squash and proposal to automate the process using machine learning is unique to this study.

Attack investigation is an important research field in forensics analysis. Many existing supervised attack investigation methods rely on well-labeled data for effective training. While the unsupervised approach based on BERT can mitigate the issues, the high degree of similarity between certain real-world attack and normal behaviors makes it challenging to accurately identify disguised attacks. This paper proposes ConLBS, an attack investigation approach that combines the contrastive learning framework and multi-layer Transformer network to realize the classification of behavior sequences. Specifically, ConLBS constructs behavior sequences describing behavior patterns from audit logs, and a novel lemmatization strategy is proposed to map the semantics to the attack pattern layer. Four different augmentation strategies are explored to enhance the differentiation between attack and normal behavior sequences. Moreover, ConLBS can perform unsupervised representation learning on unlabeled sequences, and can be trained either supervised or unsupervised depending on the availability of labeled data. The performance of ConLBS is evaluated in two public datasets. The results show that ConLBS can effectively identify at-tack behavior sequences in the cases of unlabeled data or less labeled data to realize attack investigation, and achieve the superior effectiveness compared to existing methods and models.

All kinds of vessels consist of dozens of complex machineries with rotating parts and electric motors that operate continuously in a harsh environment with excess temperature and humidity, vibration, fatigue and load. A breakdown or malfunction in one of these machineries can significantly impact the vessel’s operation and safety and consequently, the safety of the crew and the environment. To maintain operational efficiency and seaworthiness, the shipping industry invests substantial resources in preventive maintenance and repairs. This research presents the economic and technical benefits of predictive maintenance over to traditional preventive maintenance, and repair by replacement approaches in the maritime domain. By leveraging modern technology and Artificial Intelligence, we can analyze real-time operating conditions of machinery, enabling early detection of potential damages and allowing for effective planning of future maintenance and repair activities. In this paper, we propose and develop a Convolutional Neural Network that is fed with raw vibration measurements acquired in a laboratory environment from the ball bearings of a motor. Then, we investigate whether the proposed network can accurately detect the functional state of ball bearings and categorize any possible failures present, contributing to improved maintenance practices in the shipping industry.

Integrating artificial intelligence (AI) and natural language processing (NLP) technologies with project-based learning experiences in Science,Technology, Reading, Engineering, Arts, and Mathematics (STREAM) education offers the potential to enhance student engagement, critical thinking,problem-solving, and interdisciplinary learning. This paper presents a comprehensive approach to implementing AI and NLP technologies, such as ChatGPT, in STREAM project-based learning experiences, highlighting the potential advantages and challenges of this innovative educational approach. The proposed method encompasses personalized learning pathways, AI-powered research assistance, collaborative AI tools, real-time feedback, virtual mentoring, advanced simulation, and objective assessment. We also discuss the implications of this integration for educators, institutions, and students, along with recommendations for future research and development in this emerging field.

We suggest a deep learning based sensor signal processing method to remove chemical, kinetic and electrical artifacts from ion selective electrodes’ measured values. An ISE is used to investigate the concentration of a specific ion from aqueous solution, by measuring the Nernst potential along the glass membrane. However, application of ISE on a mixture of multiple ion has some problem. First problem is a chemical artifact which is called ion interference effect. Electrically charged particles interact with each other and flows through the glass membrane of different ISEs. Second problem is the kinetic artifact caused by the movement of the liquid. Water molecules collide with the glass membrane causing abnormal peak values of voltage. The last artifact is the interference of ISEs. When multiple ISEs are dipped into same solution, one electrode’s signal emission interference voltage measurement of other electrodes. Therefore, an ISE is recommended to be applied on single-ion solution, without any other sensors applied at the same time. Deep learning approach can remove both 3 artifacts at the same time. The proposed method used 5 layers of artificial neural networks to regress correct signal to remove complex artifacts with one-shot calculation. Its MAPE was less than 1.8% and R² of regression was 0.997. A randomly chosen value of AI-processed data has MAPE less than 5% (p-value 0.016).

Accurately mapping individual tree species in densely forested environments is crucial to forest inventory. When considering only RGB images, this is a challenging task for many automatic photogrammetry processes. The main reason for that is the spectral similarity between species in RGB scenes, which can be a hindrance for most automatic methods. State-of-the-art deep learning methods could be capable of identifying tree species with an attractive cost, accuracy, and computational load in RGB images. This paper presents a deep learning-based approach to detect an important multi-use species of palm trees (Mauritia flexuosa; i.e., Buriti) on aerial RGB imagery. In South-America, this palm tree is essential for many indigenous and local communities because of its characteristics. The species is also a valuable indicator of water resources, which comes as a benefit for mapping its location. The method is based on a Convolutional Neural Network (CNN) to identify and geolocate singular tree species in a high-complexity forest environment, and considers the likelihood of every pixel in the image to be recognized as a possible tree by implementing a confidence map feature extraction. This study compares the performance of the proposed method against state-of-the-art object detection networks. For this, images from a dataset composed of 1,394 airborne scenes, where 5,334 palm-trees were manually labeled, were used. The results returned a mean absolute error (MAE) of 0.75 trees and an F1-measure of 86.9%. These results are better than both Faster R-CNN and RetinaNet considering equal experiment conditions. The proposed network provided fast solutions to detect the palm trees, with a delivered image detection of 0.073 seconds and a standard deviation of 0.002 using the GPU. In conclusion, the method presented is efficient to deal with a high-density forest scenario and can accurately map the location of single species like the M flexuosa palm tree and may be useful for future frameworks.

We propose a probabilistic graphical model for discriminative substrate characterization, to support geological and biological habitat mapping in aquatic environments. The model, called a fully connected conditional random field (CRF), is demonstrated using multispectral and monospectral acoustic backscatter from heterogeneous seafloors in Patricia Bay, British Columbia, and Bedford Basin, Nova Scotia. Unlike previously proposed discriminative machine learning algorithms, the CRF model considers both the relative backscatter magnitudes of different substrates and their relative proximities. The model therefore combines the statistical flexibility of a machine learning algorithm with an inherently spatial treatment of the substrate. The CRF model predicts substrates such that nearby locations with similar backscattering characteristics are likely to be in the same substrate class. The degree of proximity and allowable backscatter similarity are controlled by parameters that are learned from the data. CRF model results were evaluated against a popular generative model known as a Gaussian Mixture model that doesn't include spatial dependencies, only covariance between substrate backscattering response over different frequencies. Both models are used in conjunction with sparse bed observations/samples in a supervised classification. A detailed accuracy assessment, including a leave-one-out cross-validation analysis, was performed using both models. Using multispectral backscatter, the GMM model trained on 50% of the bed observations resulted in a 75% and 89% average accuracies in Patricia Bay and Bedford Basin, respectively. The same metrics for the CRF model were 78% and 95%. Further, the CRF model resulted in a 91% mean cross-validation accuracy across four substrate classes at Patricia Bay, and a 99.5% mean accuracy across three substrate classes at Bedford Basin, which suggest that the CRF model generalizes extremely well to new data. This analysis also showed that the CRF model was much less sensitive to the specific number and locations of bed observations than the generative model, owing to its ability to incorporate spatial autocorrelation in substrates. The CRF approach therefore may prove to be a powerful `spatially aware' alternative to other discriminative classifiers.

As the oil industry increasingly turns to unconventional low-permeability and low-porosity reservoirs, precise post-fracture production prediction is a pivotal instrument in investment de-cision-making, formulation of energy policies, and promotion of environmental assessments. Nevertheless, despite extensive research spanning decades, the precise post-fracture production prediction based on logging parameters remains an intricate task. In this study, we gathered ex-tensive logging data and segmented the post-frac gas production during well testing on the first day to enrich the dataset. Nine pipelines were then architected using various techniques of data preprocessing, feature extraction, and advanced machine learning models. Hyperparameter op-timization was executed via the GridsearchCV. To assess the efficacy of diverse models, metrics including the coefficient of determination (R2), standard deviation (SD), and root mean square error (RMSE) were invoked. Among the several pipelines explored, the PS-NN exhibited excellent predictive capability in specific reservoir contexts (an R2 value of 0.94 and an RMSE of 48.15). In essence, integrating machine learning with logging parameters effectively assesses reservoir productivity at multi-meter formation scales. This strategy not only mitigates uncertainties en-demic to reservoir exploration but also equips petroleum engineers with monitoring on reservoir dynamics, thereby facilitating reservoir development. Additionally, this approach provides res-ervoir engineers with an efficient way of reservoir performance oversight.

Continuous greenhouse gas emissions are causing global warming and impacting the habitats of many animals. Researchers in the field of electric power are making efforts to mitigate this situation. Operating and maintaining the power grid in an economic, low-carbon, and stable is challenging. To address the issue, we propose a grid dispatching technique that combines prediction technology, reinforcement learning, and optimization technology. Prediction technology can forecast future power demand and solar power generation, while reinforcement learning and optimization technology can make charging and discharging decisions for energy storage devices based on current and future grid conditions. In the power system, the aggregation of distributed energy resources increases uncertainty, particularly due to the fluctuating generation of renewable energy. This requires the use of advanced predictive control techniques to ensure long-term economic and decarbonization goals. In this paper, we present a real-time dispatching framework that integrates deep learning-based prediction, reinforcement learning-based decision-making, and stochastic optimization techniques. The framework can rapidly adapt to target uncertainty caused by various factors in real-time data distribution and control processes. The proposed framework achieved global Champion in the NeurIPS Challenge 2022 competition and demonstrated its effectiveness in practical scenarios of intelligent building energy management.

The prediction of drug-target interactions plays a fundamental role in facilitating drug discovery, where the goal is to find prospective drug candidates. With the increase in the number of drug-protein interactions, machine learning techniques, especially deep learning methods, have become applicable for drug-target interaction discovery because they significantly reduce the required experimental workload. In this paper, we present a spontaneous formulation of the drug-target interaction prediction problem as an instance of multi-instance learning. We address the problem in three stages, first organizing given drug and target sequences into instances via a private-public mechanism, then identifying the predicted scores of all instances in the same bag, and finally combining all the predicted scores as the output prediction. A comprehensive evaluation demonstrates that the proposed method outperforms other state-of-the-art methods on three benchmark datasets.

Covid-19 has caused infections and deaths worldwide. While research in the field of Data Science has contributed good predictions of positive Covid-19 case numbers, this study's review of literature shows there is little research in the use of variants of the virus in predictions. We set out to define and evaluate novel variant features. We find that features relating to variant trends, thresholds and amino acid substitutions are especially powerful in two tasks. In the first task, predicting Covid-19 case numbers, accuracy improved from 71.53% without variant features to 82.12% with variant features. In the second task, predicting transmission severity of variants between two classes, we created a method to build some variable ensembles through selecting appropriate models that are generated with variant features. The test results showed that our ensembles are more accurate and reliable. One particular ensemble of 14 models correctly classified 90.91% of variants, outperforming other models including the popular Random Forest ensemble. In addition, as the variant features have represented more underlying information about Covid-19 pathophysiology, our ensemble methods use only a few data samples to achieve an accurate prediction. The ensemble of 14 models uses only 50 cases of each variant, an ability that could be exploited for early detection of highly infectious variants. These research findings may benefit public health professionals, policy makers, and the research community in the collective efforts to overcome this disease.

Electroencephalograms (EEGs) of children with reading disorder (RD) are characterized by higher power and coherence in slow frequencies (delta and theta bands) and lower power and coherence in the alpha band compared to EEGs of typically developed children. Neurofeedback (NFB) is useful for treating learning disorders by reinforcing the reduction of the theta/alpha ratio. This efficacy is supported by EEG power normalization and cognitive-behavioral improvement. To further explore brain changes in isolated areas, this study aimed to explore the effects of this NFB protocol on functional connectivity (coherence) in children with RD. Twenty children with RD and an abnormally high theta/alpha ratio underwent 30 sessions of NFB. After treatment, 90% of par-ticipants learned to decrease their theta/alpha ratio, and on average, children increased their reading accuracy and comprehension scores. Interhemispheric coherence diminished in the delta, theta and beta bands, mainly between frontopolar regions. Intrahemispheric coherence decreased in delta, theta and beta bands and increased in alpha band. The reduction in theta coherence be-tween the left frontal area and other regions of the left hemisphere had particular relevance for reading. These results suggest that this NFB treatment could positively impact reading-related functions in the brain networks of children with RD.

Robots for communication are developed extensively with emphasis on sympathy. This study deals with the growth of character and the control of its operation accordingly. The child has time to be alone with the nature of his/her robot friend. The child can interact with other people's emotional expressions through a robot. Step by step, the robot character will grow as the child grows. Through design studies, qualitative processes such as {customer experience audit, eye tracking, mental model diagrams, semantic differences} have been executed for the results. The participatory behavior research approach through user travel is mapped from the user's lead to the evidence-based design. This research considers how the synthetic characteristics can be applied to the physical growth of robot toys through the product design process. With the development of robot toy "Buddy", we tried making two variations on the robot to achieve recognizable growth. (1) A one-dimensional height scaling and (2) facial expression including the distance between two eyes on the screen. Observations represented children's reactions when "Buddy" was released to with the children. As an independent synthetic character, the robot was recognized by children who had the designed function. Robots for training may require more experimentation.

Through the development of the metaverse concept from the Sumerian myth (5500 - 1800 BC) and mind-altering novel, “Snow Crash” in 1992, to today’s information age, human- and society-centred urban metaverse worlds, an extension of residents and urban society where the virtual and the physically real blend and are more organically integrated, are meant to mirror the fabric of urban life with no harm to their residents. The success of urban metaverse cybercommunities depends on the quality of data-driven Smart City (SC) Digital Twins (DTs), the seamless exchange of data between cyber and physical worlds (e.g. between residents and their counterpart “Avatars’’) and the processing of the data effectively and efficiently with no vicious interventions. The potential risks in this ecosystem that incorporates Web3 can be extremer than the ones in Web2 since users are immersed with multiple tightly coupled wearable sensor-rich devices perceiving the blend of the real and the virtual with possible imminent negative experiences. This study, by analysing potential cyberthreats in the urban metaverse cyberspaces, proposes a blockchain-based Decentralised Privacy-Preserving Machine Learning (DPPML) authentication and verification technique, which uses the metaverse immersive devices and can be instrumented effectively against identity impersonation and theft of credentials, identity, or avatars.

Microsoft has implemented several measures to defend against macro viruses, including the use of Anti-malware Scan Interface (AMSI) and automatic macro blocking. Nevertheless, evidence shows that threat actors have found ways to bypass these mechanisms. As a result, phishing emails continue to apply malicious macros as their essential attack vector. In this paper, we analyze 77 obfuscation features from the attacker’s point of view and extract 46 suspicious keywords in macro. We first combine the above two types of features to train machine learning models on a public dataset. Then, we carry out the same experiment on self-constructed dataset, a collection of newly discovered samples, to see if our proposed method could discover the unseen malicious macros. Experimental results demonstrate that, comparing with the existing researches, our proposed method has a higher detection rate and better consistency. Furthermore, ensemble multi-classifiers with distinct feature selection further improve the detection performance.

This paper proposes a multimodal fusion 3D target detection algorithm based on the attention mechanism to improve the performance of 3D target detection. The algorithm utilizes point cloud data and information from camera. For image feature extraction, the ResNet50+FPN architecture extracts features at four levels. Point cloud feature extraction employs the voxel method and FCN to extract point and voxel features. The fusion of image and point cloud features is achieved through regional point fusion and regional voxel fusion methods. After information fusion, the Coordinate attention mechanism and SimAM attention mechanism extract fusion features at a deep level. The algorithm's performance is evaluated using the DAIR-V2X dataset. The results show that compared to the Part-A2 algorithm, the proposed algorithm improves the mAP value by 7.9% in BEV view and 7.8% in 3D view at IOU=0.5 (cars) and IOU=0.25 (pedestrians and cyclist). At IOU=0.7 (cars) and IOU=0.5 (pedestrians and cyclist), the mAP value of the SECOND algorithm is improved by 5.4% in the BEV view and 4.3% in the 3D view, compared to other comparison algorithms.

The creation of crop-type maps from satellite data has proven challenging, often impeded by a lack of accurate in-situ data. This paper aims to demonstrate a method for crop-type (ie. Maize, Wheat and Other) recognition based on Convolutional Neural Networks using a bottom-up approach. We trained the model with a highly accurate dataset of crowdsourced labelled street-level imagery. Classification results achieved an AUC of 0.87 for wheat, 0.85 for maize and 0.73 for other. Given that wheat and maize are the two most common food crops globally, combined with an ever-increasing amount of available street-level imagery, this approach could help address the need for improved crop-type monitoring globally. Challenges remain in addressing the noisy aspect of street-level imagery (ie. buildings, hedgerows, automobiles, etc.), where a variety of different objects tend to restrict the view and confound the algorithms

Magnesium matrix composites have attracted significant attention due to their lightweight nature and impressive mechanical properties. However, the fabrication process for these alloy compo-sites is often time-consuming, expensive, and labor-intensive. To overcome these challenges, this study employed machine learning (ML) techniques to predict the mechanical properties of magnesium matrix composites. Regression models were utilized to forecast the yield strength of magnesium alloy composites reinforced with various materials. The study incorporated previous research on matrix type, reinforcement type, heat treatment, and mechanical working. The re-gression models employed in this study included decision tree regression, random forest re-gression, extra tree regression, and XGBoost regression. Model performance was assessed using metrics such as RMSE and R2. The XGBoost Regression model out-performed others, exhibiting an R2 value of 0.94 and the lowest error rate. Feature importance analysis indicated that the rein-forcement particle form had the greatest influence on the mechanical properties. The study iden-tified the optimized parameters for achieving the highest yield strength, which was 186.99 MPa. Overall, this study successfully demonstrates the effectiveness of ML as a valuable tool for opti-mizing the production parameters of magnesium matrix composites.

Buildings are responsible for approximately 40% of the world’s energy consumption and 36% of the total carbon dioxide emissions. Building occupancy is essential, enabling Occupant-Centric Control for zero emissions and decarbonization. Although existing machine learning and deep learning methods for building occupancy prediction have achieved remarked progress, their analyses remain limited when applied to complex real-world scenarios. Besides, there is a high expectation for Transformer algorithms to predict building occupancy accurately. Therefore, this paper presents an Occupancy Prediction Transformer network (OPTnet). We fuse and feed multi-sensor data (building occupancy, indoor environmental conditions, HVAC operations) into a Transformer model to forecast the future occupancy presence in multiple zones. We perform experimental analysis and compare it to different occupancy prediction methods (e.g., Decision Tree, Long Short-Term Memory networks, Multi-layer perceptron) and diverse time horizons (1,2,3,5,10,20,30mins). The performance metrics (e.g., Accuracy and Mean Squared Error) are employed to evaluate the effectiveness of prediction algorithms. Our OPTnet method achieves superior performance on our experimental two-week data compared to existing methods. The improved performance signifies its potential to enhance HVAC control systems and energy optimization strategies. We will make the code publicly available at https://github.com/kailaisun/occupancy-prediction-binary to promote transparency and reproducibility.

Previously, we described the initial use of Flipped Team‐Based learning (FTBL) defined as TBL approach combined with flipped classroom learning methodology, in which students previewed online lectures and applied their knowledge in different in-class activities. The purpose of the present study is to review the progress within this approach and to investigate how constant changes in team allocation can affect student’s perception regarding this modified FTBL approach. Although students showed reluctance initially to get out of their ‘comfort zone’, our findings show that learners perceived the adoption of the continued random allocation, and became accustomed to this learning approach, which finally assisted them to enhance their team-work skills and classroom performance, to develop their reflective capabilities as well as improving their rapport building skills, learning and academic performance. Learners also believed that this learning strategy that creates critical incidents can simulate their future work environment as they might be expected to work in unfamiliar situations. Therefore, the present study indicated strong support for the modified FTBL method and was seen to work exceptionally well, despite some minor problems that students can experience working in a team and/or with different teammates in every session.