Emotions play a vital role in understanding human behavior and interpersonal relationships. The ability to recognize emotions through Electroencephalogram (EEG) signals offers an alternative to traditional methods, such as questionnaires, enabling the identification of emotional states in a non-intrusive manner. Automatic emotion recognition holds great potential, eliminating the need for clinical examinations or physical visits, thereby contributing significantly to the advancement of Brain-Computer Interface (BCI) technology. However, one of the key challenges lies in effectively selecting and extracting relevant features from the EEG signal to establish meaningful distinctions between different emotional states. The process of feature selection is often time-consuming and demanding. In this research, we propose a groundbreaking approach for automatically identifying three emotional states (positive, negative, and neutral) by leveraging auditory stimulation of EEG signals. Our novel method directly applies the raw EEG signal to a Convolutional Neural Network-Long Short Term Memory (CNN-LSTM) architecture, bypassing the conventional feature extraction and selection steps. This unconventional approach offers a significant departure from existing literature. Our proposed network architecture comprises ten convolutional layers, followed by three LSTM layers and two fully connected layers. Through extensive simulations and evaluations on 12 active channels, our algorithm demonstrates exceptional performance, achieving an accuracy of 97.42\% and 95.23\% for the binary classification of negative and positive emotions, as well as a Cohen's Kappa coefficient of 0.96 and 0.93 for the three-class classification (negative, neutral, and positive), respectively. These promising results highlight the efficacy of our novel methodology and its potential implications in advancing emotion recognition using EEG signals.

Dynamic neural networks (DNN) are types of artificial neural networks (ANN) that are designed to work with sequential data where context in time is important. In contrast to traditional static neural networks that process data in a fixed order, dynamic neural networks use information about past inputs, which is important if dynamic of a certain process is emphasized. They are widely used in natural language processing, speech recognition and time series prediction. In industrial processes, their use is interesting for the prediction of difficult-to-measure process variables. In an industrial process of isomerization, it is crucial to measure the quality attributes affecting the octane number of gasoline. Process analyzers that are commonly used for this purpose are expensive and subject to failures, therefore, in order to achieve continuous production in case of malfunction, mathematical models for estimating the product quality attributes are imposed as a solution. In this paper, mathematical models were developed using dynamic recurrent neural networks (RNN), i.e., their subtype of a long short-term memory (LSTM) architecture. The results of the developed models were compared with the results of several types of other data-driven models developed for an isomerization process, such as multilayer perceptron (MLP) artificial neural networks, support vector machines (SVM) and dynamic polynomial models. The obtained results are satisfactory, which suggests a good possibility of application.

The deformation monitoring information of concrete dams contains some high-frequency com-ponents, and the high-frequency components are strongly nonlinear, which reduces the accuracy of dam deformation prediction. In order to solve such problems, this paper proposes a concrete dam deformation monitoring model based on empirical mode decomposition (EMD) combined with wavelet threshold noise reduction and sparrow search algorithm (SSA) optimization of long short-term memory network (LSTM). The model uses EMD combined with wavelet threshold to decompose and denoise the measured deformation data. On this basis, the LSTM model based on SSA optimization is used to mine the nonlinear function relationship between the reconstructed monitoring data and various influencing factors. The example analysis shows that the model has good calculation speed, fitting and prediction accuracy and it can effectively mine the date char-acteristics inherent in the measured deformation, and reduce the influence of noise components on the modeling accuracy.

Muscarinic acetylcholine receptors are well-known for their crucial involvement in hippocampus-dependent learning and memory, but the exact roles of the various receptor subtypes (M1-M5) are still not fully understood. Here, we studied how M1 and M3 receptors affect plasticity at the mossy fiber (MF)-CA3 pyramidal cell synapse. In hippocampal slices from M1/M3 receptor double knockout (M1/M3-dKO) mice, the signature short-term plasticity of the MF-CA3 synapse was not significantly affected. However, the rather unique, NMDA receptor-independent and presynaptic form of long-term potentiation (LTP) of this synapse was much larger in M1/M3-deficient slices compared to wild type slices, in both field potential and whole-cell recordings. Consistent with its presynaptic origin, induction of MF-LTP strongly enhanced the excitatory drive onto single CA3 pyramidal cells, with the effect being more pronounced in M1/M3-dKO cells. In an earlier study [1], we found that deletion of M2 receptors in mice disinhibits MF-LTP in a similar fashion, suggesting that endogenous acetylcholine employs both M1/M3 and M2 receptors to constrain MF-LTP. Importantly, such synergism was not observed for MF long-term depression (LTD). Low-frequency stimulation, which reliably induced LTD of MF synapses in control slices, failed to do so in M1/M3-dKO slices and gave rise to LTP instead. In striking contrast, loss of M2 augmented LTD when compared to control slices. Taken together, our data demonstrate convergence of M1/M3 and M2 receptors on MF-LTP, but functional divergence on MF-LTD, the net effect being well-balanced bidirectional plasticity of the MF-CA3 pyramidal cell synapse.

Arable Сhernozems with high SOC contents have the potential to be significant sources of GHGs, and climate change is likely to increase SOC losses, making the issue of carbon sequestration in this region even more important. The prospect of maintaining SOC stock or increasing it by 4% an-nually under planned management practice modifications for the period up to 2090 was evaluated using a long-term experiment on Haplic Chernozem in the Rostov Region, Russia. In this study, we used the RothC model to evaluate SOC dynamics for three treatments with mineral and organic fertilization under two adaptation scenarios vs. business as usual, as well as under two climate change scenarios. Correction of crop rotation and the application of organic fertilizers at high rates are essential tools for maintaining and increasing SOC stocks. This can maintain SOC stock at the level of 84–87 Mg∙ha-1 until the middle of the 21st century, as the first half of the century is con-sidered the most promising period for the introduction of adaptation measures for the additional accumulation of SOC on Chernozems. Part of the additional accumulated SOC is expected to be lost before 2090.

Evaluation of the patients after the surgical correction of idiopathic scoliosis in long-term follow-up with the clinical neurophysiology methods has not been presented in detail. This study aimed to compare the results of the neurophysiological studies in 45 girls with scoliosis of Lenke 1-3 types performed pre- (T0), postoperatively (T1, one week after surgery), and 6 months after the surgery (T2). Parameters of surface electromyography during the attempt of maximal contraction (mcsEMG) and transcranially evoked motor evoked potentials (MEP) recorded from anterior tibial muscles, as well as the electroneurography (ENG) of peripheral transmission in the peroneal nerve motor fibers were compared. The results indicate that efferent neural conduction function both centrally and peripherally as well as TA muscle function improve immediately after surgical correction of scoliosis (at p=0.05) and further normalization appears after six months in long-term follow-up (at p=0.03). It has been found in sEMG recordings, that the TA muscle motor unit recruitment function after half a year from surgical treatment in IS patients is comparable to the normal condition. ENG recordings results indicated the gradual retreat of the motor fibers injury symptoms mainly of the axonal type in peroneal nerves, the surgeries improved also the lumbar ventral roots’ neural transmission to the functional status considered as normal. MEP amplitude parameters recorded after the surgical scoliosis correction in T1 indicated a slight improvement in the efferent transmission of neural impulses within the fibers of the spinal tracts; in long-lasting T2 observation, they reached values comparable to those recorded in healthy volunteers bilaterally. Preoperatively (T0) the results of all neurophysiological study parameters in IS patients were asymmetrical at p=0.036-0.05 and recorded as worse on the concave side, suggesting the lateralization of neurological motor deficits. One week postoperatively (T1), this asymmetry was recorded as gradually reduced, showing almost no difference between the right and left sides six months later (T2). The presented algorithm of the diagnostic proceeding pre-, early post-, and long-lasting postoperatively using mcsEMG, MEP, and ENG neurophysiological examinations can be significant not only in making the final decision regarding surgical treatment and its personalization but also helps in precise ascertaining its effects as well as in predicting the final result of IS treatment.

This research work deals with the formulation of numerical scheme with algorithm to compute singularly perturbed partial differential equations(SP-PDEs) involving a significant negative shift. The problem involves small parameter which causes a rapidly changing boundary layers in the vicinity of a body, and the negative shift term causes interior layer. Such appearance of an abruptly varying layers causes difficulties to find the exact solution and it is not adequate to employ classical numerical methods. The technique presented in this research work simplifies these challenges and yields accurate numerical solution.The stability and convergence analyses of the methods are examined and proven. To test the developed technique, numerical experiments are carried out and confirmed with theoretical analysis.

The results of numerical experiments with the chemistry-climate model of the Institute of Numerical Mathematics of the Russian Academy of Sciences - Russian State Hydro-Meteorological University (INM-RSHU CCM) in comparison with the MERRA2 reanalysis data are presented. To assess the impact of sea surface temperature (SST) on stratospheric processes and the ozone layer, numerical experiments were carried out for scenarios corresponding to the El Niño and La Niña Southern Oscillation phases. It is shown that the El Niño phase contributes to the enhancement of heat and mass transfer to the polar stratosphere, which strengthen the Brewer-Dobson circulation and the flux of wave activity that affect the zonal wind. As a result, during El Niño, the probability of sudden stratospheric warming increases and the stratospheric polar vortex (SPV) weakens, re-sulting in an increase in total polar stratospheric ozone. The La Niña phase, on the contrary, is accompanied by a weakening of heat and mass transfer from the troposphere to the stratosphere, as a result of which the stratospheric polar vortex strengthens and the total column ozone decreases. It is shown that the effect associated with the Southern Oscillation is more pronounced in the Northern Hemisphere than in the Southern Hemisphere. Numerical experiments were also carried out for scenarios in which the sea surface temperature was set corresponding to the early 1980s, and the results were compared with calculations in which the SST was set corresponding to the late 2010s. This allowed us to assess the influence of the multiyear trend in the SST increase on stra-tospheric processes and the ozone layer. It was shown that the SST increase is most pronounced in the Arctic, which corresponds to the phenomenon of "Arctic amplification", and therefore, the heat and mass transfer to the stratosphere increases in the Northern Hemisphere, which contributes to some weakening of the SPV and increase of the total column ozone. But, compared to the Southern Oscillation, the influence of the SST trend on the SPV and total column ozone is much weaker and is observed mainly in the Northern Hemisphere.

The intense motion of a ship can greatly impacts the comfort of crew members and the safety of the vessel. Therefore, accurately estimating and predicting ship attitudes has become an important issue. This paper introduces the latest development in functional deep learning model called DeepOnet. It takes wave height as input and ship motion as output, using a cause-to-result prediction approach. The modeling data used in this study is sourced from publicly available experimental data from the Iowa Institute of Hydraulic Research. Firstly, parameters system tuning was conducted for the neural network's hyperparameters to determine the appropriate combination of parameters. Secondly, the DeepOnet model for wave height and multi-degree-of-freedom motion was established, and the influence of increasing time steps on prediction accuracy was examined. Finally, a comparison was made between the DeepOnet model and the classical time series model LSTM. It was found that the DeepOnet model had a 10-fold improvement in accuracy for roll and heave attitudes. Moreover, as the forecast duration increased, the advantage of DeepOnet showed a trend of strengthening. As a functional prediction model, DeepOnet provides a new promising tool for very short-term prediction of ship motion.

Deep Learning (DL) oriented document processing is widely used in different fields for extraction, recognition, and classification processes from raw corpora of data. The article examines the application of deep learning approaches, based on different neural network methods, including Gated Recurrent Unit (GRU), Long Short-Term Memory (LSTM), and Convolutional Neural Networks (CNN). The compared models were combined with two different word embedding techniques, namely: Bidirectional Encoder Representations from Transformers BERT and Gensim Word2Vec. The models are designed to evaluate the performance of architectures based on neural network techniques for the classification of CVs of Moroccan engineering students at ENSAK(National School of Applied Sciences of Kenitra, Ibn Tofail University). The used dataset included resumes collected from engineering students at ENSAK in 2023 for a project on the employability of Moroccan engineers in which new approaches were applied, especially machine learning, deep learning, and big data. Accordingly, 867 resumes were collected from five specialties of study (Electrical Engineering, Networks and Systems Telecommunications, Computer Engineering, Automotive Mechatronics Engineering, Industrial Engineering). The results revealed good performance of the proposed models based on the BERT embedding approach compared to models based on the Gensim Word2Vec embedding approach. Accordingly, the CNN-GRU/Bert model achieved slightly better accuracy with 0.9251 compared to other hybrid models.

Zinc is an important nutrient for plant growth and development. Its availability is influenced by zinc solubilizing microbes (ZSM). Effects of commonly promoted agronomic practices on the abundance of ZSM are so far not well understood. In this study, we assessed the effects of farmyard manure (FYM) application, either sole, or in combination with residue and/or inorganic fertilizer inputs on ZSM community structure using 11 treatments in a long-term (17 years) integrated soil fertility management experiment located in western Kenya, The study was conducted in 2019. Bacterial and fungal community composition (at 0-15 cm depth) was evaluated by amplicon sequencing on an Illumina Miseq platform following the manufacturer’s guidelines. Microbial diversity indices (Shannon, Simpson and Chao1) and statistical analyses were computed using R project packages. Putative ZSM (i.e., the ZSM generally considered to possess the zinc solubilizing capabilities) were clustered in two major clades based on either application or no application of FYM. Sole application of FYM significantly (P<0.05) increased the abundance of several ZSM under maize-tephrosia rotation. In addition, systems with combined application of FYM with other inputs generally showed significantly increasing trends for some ZSM under maize-tephrosia rotation. Moreover, combined application of FYM and P rather than P only significantly increased abundance of some ZSM under maize-monocropping systems. Furthermore, besides affecting ZSM abundance, soil chemical variables involving soil organic carbon (SOC), total N and Olsen P significantly increased with FYM application. This study indicated that management practices like application of FYM that increase SOC, and other soil chemical parameters, also/concomitantly increase ZSM abundance; implying enhanced capacities for microbial-linked zinc availability.

Introduction: Congenital heart defects (CHD) are the most frequent group of major congenital anomalies, accounting for almost 1% of all births. They comprise a very heterogeneous group of birth defects in terms of their severity, clinical management, epidemiology and embryologic origins. Taking this heterogeneity into account is an important imperative to provide reliable prognostic information to patients and their caregivers, as well as to compare results between centers or to assess alternative diagnostic and treatment strategies. The Anatomic and Clinical Classification of CHD (ACC-CHD) aims to facilitate both the CHD coding process and data analysis in clinical and epidemiological studies. Objectives: 1) To describe the long-term childhood survival of newborns with CHD and 2) To develop and validate predictive models of infant mortality based on the ACC-CHD. Materials and Methods: This study is based on data from a population-based, prospective cohort study: Epidemiological Study of Children with Congenital Heart Defects (EPICARD). The final study population comprised 1881 newborns with CHD after excluding cases that were associated with chromosomal and other anomalies. Statistical analysis included non-parametric survival analysis and flexible parametric survival models. The predictive performance of models was assessed by Harrell’s C index and the Royston-Sauerbrei R2 with internal validation by bootstrap. Results: The overall 8-year survival rate for newborns with isolated CHD was 0.96 [0.93- 0.95]. There was a substantial difference between the survival rate of the categories of ACC-CHD. The highest and lowest 8-year survival rates were 0.995 [0.989-0.997] and 0.34 [0.21-0.50] for “interatrial communication abnormalities and ventricular septal defects” and "functionally univentricular heart", respectively. Model discrimination as measured by Harrell’s C was 87% and 89% for the model with ACC-CHD alone and the full model, which included other known predictors of infant mortality, respectively. The predictive performance as measured by R2 was 45% and 50%, for the ACC-CHD alone and the full model. These measures were essentially the same after internal validation by bootstrap. Conclusion: The ACC-CHD classification provided the basis of a highly discriminant survival model with good predictive ability for the 8-year survival of newborns with CHD. Prediction of individual outcomes remains an important clinical and statistical challenge.

We study the short−term memory capacity of ancient readers of the original New Testament written in Greek, of its translations to Latin and modern languages. To model the short–term capacity, we have considered the number of words per interpunctions, the “word interval” , because this parameter can model how the human mind memorizes “chunks” of information. Since can be calculated for any alphabetical text, we can perform experiments − otherwise impossible − with ancient readers by studying the literary works they used to read. The “experiments” compare the of texts of a language/translation to those of another language/translation by measuring the minimum average probability of finding joint readers (those who can read both texts because of their similar short–-term memory capacity) and by defining an “overlap index”. We also define a population of universal readers who can read any the New Testament written in alphabetical language. More than 50% of the readers of specific languages overlap with the universal readers with probability . Future work is vast, with many research tracks, because alphabetical literatures are very large and allow many experiments, such as comparing authors, translations or even texts written by artificial intelligence tools.

The present study examines the long-term care service awareness, needs, and usage intention of elderly male homosexuals in Taiwan and their ideal long-term care service model. This study of five elderly male homosexual subjects aged 66 to 73 years was intended as a preliminary exploration. Interviews were used to collect data. The study determined that the five subjects had high awareness of long-term care services, as they had actual experience of long-term care services or even participated in care service staff training to obtain certificates. Some even had experience in applying for home care services and experienced problems during use. The subjects perceived that they were very likely to require long-term care services in the future and tended to use home care services if they required long-term care services. Due to their personal experiences, the subjects had negative awareness of long-term care services and worried that long-term care service staff had poor attitudes toward homosexuals. The subjects considered the most important aspects of long-term care to be basic medical care and lifestyle care. However, they worried that long-term care staff would delay or refuse to provide such services due to the subjects’ sexual orientation or stereotypes, and they were concerned above all about the “friendly attitude” of long-term care staff. They hoped that long-term care staff were friendly toward homosexuals and did not discriminate against them, and they felt that it would be more appropriate for homosexual long-term care staff to provide assistance. In terms of vision, while preferring organizations with homosexual employees, the subjects worried that they would be stigmatized and discriminated against . With regard to ideal long-term care services, while considering institutions with homosexual staff to be ideal, the subjects also worried that these would be labeled as institutions that were dedicated to homosexuals, potentially resulting in discrimination. Therefore, they hoped that the sexuality sensitivity of long-term care staff could be improved and that they would undergo professional continuing education to learn about homosexuals, the situation and care needs of homosexuals, and care techniques for homosexuals.

The physiological process of iodine uptake in the thyroid is used for 131I treatment of thyroid diseases. Children are more sensitive to radiation compared to adults and may react differently to 131I exposure. The aims of this study were to evaluate the effects on thyroid protein expression in young and adult rats one year after 131I injection, and identify potential biomarkers related to 131I exposure, absorbed dose, and age. Twelve Sprague Dawley Rats (young and adults) were i.v. injected with 50 kBq or 500 kBq 131I and were killed twelve months later. Twelve untreated rats were used as age-matched controls. Quantitative proteomics, statistical analysis and evaluation of biological effects was performed. The effects of irradiation were most prominent in young rats. Protein biomarker candidates were proposed related to age, absorbed dose, thyroid function and cancer and a panel was proposed for 131I exposure. In conclusion, the proteome of rat thyroid was differentially regulated twelve months after low-intermediate dose exposure to 131I in both young and adult rats. Several biomarker candidates are proposed for 131I exposure, age, and many of them are known to be related to thyroid function or thyroid cancer. Further research on human samples are needed for validation.

In-wheel motor vehicles are gaining attention as a new type of electric vehicle due to their efficient power units located inside each wheel hub. However, they are more susceptible to wheel resonance due to the increase in unsprung mass caused by the motor's weight. This can result in decreased both the ride comfort and driving stability. To resolve this issue, this study aims to apply an optimal switching controller with a semi-active n actuator; magnetorheological (MR) damper. For the implementation of the optimal switching controller, a road type classification is also carried out. An acceleration sensor is used for road type classification, and control logics include a ride comfort controller; the linear quadratic regulator (LQR _Paved Road) and a wheel motion controller (LQR_Off Road) for improved driving stability. For paved roads, the LQR_Paved Road control input is applied to the MR damper. However, if a road type prone to wheel resonance is detected, the control logic switches to the LQR_Off Road. During the transition, a weighted average of both LQR_ Paved Road and LQR_Off Road control input is applied to the actuator. Computer simulations are undertaken to evaluate vibration control performance, including the ride comfort and driving stability at various road profiles.

Background: PFO (Patent foramen ovale) is a common defect that affects about 25% of the population. Although its presence is asymptomatic in the majority of the cases, the remaining part becomes overt with different symptoms, including cryptogenic stroke. The PFO closure procedure is widely available to date with the most used Amplatzer PFO Occluder, also in complex anatomy, but the performance of another device, the GORE Septal Occluder (GSO), has not been completely explored with regard to different septal anatomies. Methods: From March 2012 to June 2020, 118 consecutive patients with an indication for PFO closure were treated using the GSO system and included in a prospective analysis and followed. After 12 months, every patient underwent transcranial Doppler to evaluate the effectiveness of treatment. Results: of 111 patients evaluated, 107 showed effective PFO closure (96,4%) and 4 showed a residual shunt (3,6%). To better evaluate the device performance, the overall population was sorted into 2 clusters based on the echocardiographic characteristics. The main difference between groups was for PFO width (4,85 ± 1,8 vs 2,9 ± 1 mm, p <0,001) and PFO tunnel length (12,6 ± 3,8 vs 7,2 ± 2 respectively, p <0,001), allowing identification of complex anatomy and simple anatomy, respectively. Regardless of the aforementioned cluster, the GSO performance to reach an effective closure was independent of anatomy type and the chosen device size. Conclusion: GSO device showed efficacy and safety at 1-year follow-up in patients with at least 1 anatomical factor of complexity of PFO, irrespectively from the level of complexity itself.

Combined sewer overflow (CSO) is a significant environmental concern and public health (e.g., water contamination, eutrophication, and beach closure). The Environmental Protection Agency (EPA) has introduced the National Pollutant Discharge Elimination System (NPDES) permitting program to regulate and address this matter. This program mandates the control of CSOs for more than 700 municipalities obligated to devise Long-term Control Plans (LTCPs) to curb combined sewer overflows and bring them down to safe levels. LTCP involves diverse strategies, including sewer separation, green infrastructure improvements, and conventional gray infrastructure upgrades. This study investigates several municipalities’ solutions for CSO problems in conventional methods and wireless sensor technology as real-time control. The investigation mainly focuses on a comparative analysis of two cities, Richmond, Virginia, and South Bend, Indiana, such as average rainfall, the frequency of overflows, and the capacity of treatment plants. The findings indicate that integrating sensor technology could significantly enhance modeling endeavors, bolster the capacity of existing structures, and substantially enhance preparedness for storm events. The EPA’s Storm Water Management Modeling (SWMM) software is utilized. Through an analysis of SWMM data, the study suggests the potential for leveraging wireless sensor technology to achieve more robust control over CSOs as a part of LTCPs.

Data-driven models (DDMs) are extensively used in environmental modeling but face challenges due to limited training data and potential results not adhering to physical laws. To address this challenge, this study developed a process-guided deep learning (PGDL) model, integrating a long short-term memory (LSTM) neural network and a process-based model (PBM), CE-QUAL-W2 (W2), to predict water temperature in a stratified reservoir. The PGDL included an energy constraint term from W2's thermal energy equilibrium into the cost function of the LSTM, besides the mean square error term. In PGDL, parameters were optimized by penalizing deviations from the energy law, ensuring adherence to physical constraints. Compared to LSTM, PGDL demonstrated enhanced satisfaction with the energy balance and superior performance in water temperature prediction. Even with less field data for training, PGDL outperformed both LSTM and calibrated W2 after pre-training with data generated using the uncalibrated W2. Therefore, integration of DDM with a PBM ensured physical consistency in water temperature prediction for complex stratified reservoirs with limited data. Moreover, pre-training the PGDL with PBM proved highly effective in mitigating bias and variance due to insufficient field measurement data.

Arctic sea ice prediction holds significant importance for facilitating Arctic route planning, optimizing fisheries management, and advancing the field of sea ice dynamics research. While various deep learning models have been developed for sea ice prediction, they predominantly operate at the seasonal or sub-seasonal scale, often focusing on localized areas, and few cater to full-region daily scale prediction. This study introduces the use of spatiotemporal sequence data prediction models, namely, the convolutional LSTM (ConvLSTM) and predictive recurrent neural network (PredRNN), for the prediction of sea ice concentration (SIC). Our analysis reveals that, when solely utilizing SIC historical data as the input, the ConvLSTM model outperforms the PredRNN model in SIC prediction. To enhance the model's capacity to capture spatiotemporal relationships between multiple variables, we expanded the range of input data types to form the ConvLSTM-multi and PredRNN-multi models. Experimental findings demonstrate that the ConvLSTM-multi model excels in assimilating the influence of reanalysis data on sea ice within the sea ice edge region, thus exhibiting superior performance in predicting daily Arctic SIC over the subsequent 10 days. Furthermore, sensitivity tests on various model parameters highlight the substantial impact of sea surface temperature and prediction date on the accuracy of daily sea ice prediction. Additionally, meteorological and oceanographic parameters primarily affect the prediction accuracy of the thin ice region at the edge of the sea ice.

Humeral resurfacing arthroplasty (HRA) entails the substitution of the articular surface alone with a prosthetic cap without stem. It is a more conservative procedure which can be easily converted in a total arthroplasty if necessary. The present study aimed to evaluate the clinical and radiographical outcomes in a series of patients treated with HRA. 33 patients with a mean fol-low-up of 11 years were clinically (Constant score; Disability of the Arm, Shoulder and Hand score, DASH) and radiographically assessed before and after surgery. Constant and DASH score improved significantly after surgery, and only 2 cases needed revision surgery. HRA represents a valid therapeutic option in selected cases to improve the quality of life and delaying the need for more invasive procedures.

This paper has a twofold objective. The first aim is to present a comprehensive bibliographical analysis of journal articles and book chapters on fuzzy-random option pricing (FROP) over the WoS and SCOPUS databases. It follows PRISMA criteria and takes special care of developments in continuous time. Thus, we described the principal findings about the research streams, outlets and authors of this topic and lets us suggest further research. The second contribution is motivated by the fact that the bibliographical revision has identified a lack of developments on equilibrium models on the yield curve. This question motivates extending Vasicek’s yield curve equilibrium model to introduce fuzziness in the parameters that govern interest rate movements (speed of reversion, equilibrium short-term interest rate, and volatility). Likewise, this paper develops an empirical application on the term structure of fixed income public bonds with the highest credit rating in the Euro area.

Objective: The long-term care insurance (LTCI) system was introduced in 2000 to provide financial support for long-term care services to older adults. What is known on this topic report focuses on a comparative analysis of how the LTCI system has affected older people in Japan. Methods: We obtained data from the literature about the LTCI system, compared and analysed the collected articles, and concluded. Results: The LTCI system has improved access to long-term care services, with more older adults receiving care than before the system’s implementation. This service has improved the quality of life for many elderly individuals who require assistance with daily activities. Furthermore, the LTCI system has alleviated the financial burden of long-term care services since the insurance covers a large portion of the costs. It has made older adults maintain their independence and dignity, as they can afford the necessary care without relying solely on their savings or burdening their families. Conclusions: Japan’s LTCI system has positively impacted older adults by improving access to long-term care services and alleviating financial burdens. However, challenges such as increasing demand and variations in service quality need to be addressed to ensure the system effectively supports older people in Japan.

The rapid ageing of populations around the World is creating complex challenges for national governments. The establishment of sustainable and equitable long-term care systems for old and dependent people is one of the main issues of social policy in developed countries. The aim of this work is to define a cost model for residential and day care centres for dependent persons in Cantabria (Spain). The cost model will make it possible to establish the theoretical cost of attending to the needs of the different types of dependent persons in the different types of care centres, and the methodology used could be extrapolated to other regions. The daily cost per user for elderly residential care is €53.72. The cost per user in elderly day centres (5 days) is 32.56 euros. In residential centres for people with disabilities, the values range between €47.41 and €75.25 depending on the category of the centre. In three categories of centres the public price is not enough to cover the cost (physical disability, intellectual disability, mental illness – low care), and therefore the administration should reconsider their public prices for these kind of centres if they want to really contribute to the sustainability of these residential care centres. This research will have important implications for policy-makers in a context of fulfilment of SDGs and where better support for old and disabled people and their carers, as well as fair and efficient financing of social care services, are essential to address the current and future challenges of dependency.

With the development of sensor and communication technologies, Internet of Things (IoT) subsystem is gradually becoming a crucial part in vehicles. It can effectively enhance functionalities of vehicles. However, new attack types are also emerged. For example, a driver with the smart key in the pocket can push the start button to start a car when. At the same time, security issues in the push-to-start scenario are alerted, such as smart key forgery. In this paper, we propose a vehicle Passive Entry Passive Start (PEPS) system that uses deep learning algorithms to recognize the driver by the Electrocardiogram (ECG) signals measured by his or her smart watch. ECG signals are used for verification. Smart watches, as a new smart key of PEPS system, can replace traditional smart keys to improve security. Experiment results show that Long Short-Term Memory (LSTM) models have achieved the best accuracy score for identity recognition (91%) when single ECG cycle is used. However, it takes at least 30 minutes for training. The training time of Auto Encoder successfully reduces to 5 minutes. When 15 continuous ECG cycles are used, it can achieve 100% identity accuracy.

Genetic parameters of overreach, term and acuteness in Peruvian Paso horses (PPH) have not been determined to date. It is important to estimate these parameters for application in PPH breeding, therefore, the aim of this study was to estimate the heritability, repeatability, and genetic correlations under field conditions of overreach, term, and acuteness of PPHs. The study included records of up to 137 stallions and mares. All measurements were recorded in MP4 video format with a resolution of 1920 x 1080 megapixels and at 60 frames per second. All traits were measured three times (once per stride), and each trait was analyzed. KINOVEA software version 0.9.5 was used to analyze the measurements. A multivariate repeated measures animal model with sex effect was used to estimate the variance components for each trait using WOMBAT software. The results showed heritability of 0.411, 0.476 and 0.405, for the traits of overreach, term, and acuteness, respectively. Repeatability was high in all traits (&gt; 0.78). Genetic additive correlations ranged from -0.30 to 0.49. It can be concluded that overreach and term have high heritability values, which allows these traits to respond better in a selection process, unlike acuteness which has a moderate heritability value.

Objective: To examine variations in end-of-life care for breast cancer, heart failure, and dementia patients. Data and methods: Data from four Norwegian health registries were linked using a personal identification number. Longitudinal trends over 365 days and the type of care on the final day of life were analyzed using descriptive techniques and logistic regression analysis. Results: Patients with dementia were most commonly placed in nursing homes, while patients with heart failure and breast cancer were more frequently hospitalized. Breast cancer and heart failure patients had a higher likelihood of dying at home. The presence of general practitioners increased the probability of home-based end-of-life care for cancer patients, while non-physician healthcare workers increased the likelihood of home-based care for the other patient groups. Conclusion: Diagnoses, individual characteristics, and service availability are all associated with the place of death in end-of-life care.

Pure silver as well as silver compounds have been used as antiseptics for centuries. Nowadays, the most frequently applied form of silver is a nanoparticle form. Several studies demonstrate the ability of silver nanoparticles and silver salts to accumulate in the brain of laboratory ani-mals as well as the ability of silver nanoparticles to influence behavioral and cognitive functions of laboratory mammals. For the first time, the influence of a silver salt, in particular, silver citrate on the change dynamics of behavioral and cognitive functions of laboratory mice at the daily oral exposure during 30, 60, 120 and 180 days is investigated in the present research. We observed the anxiety increase after 30 days of the exposure as well as the ability of mice to adapt to the prolonged oral exposure of silver citrate manifested in the anxiety decrease, increase of locomotor activity and the tendency of long-term contextual memory improvement. Increase of locomotor activity and memory improvement might be associated with the sensitivity increase and regarded as compensatory mechanism to anxiety increase. The observed phenomenon is classified as a hormetic effect. A comparison with the analogous study implemented polyvi-nylpyrrolidone coated silver nanoparticles is carried out. It was concluded that the nanoform is more neurotoxic.

Recently, various applications of indoor location-based services using Wi-Fi fingerprint data have been studied. However, since fingerprint data collected at a specific location does not have continuous information between signals and correlation information between signals, path data cannot exist in principle. Therefore, most indoor positioning technologies are based on predicting the position of a stationary positioner at a specific location based on signal data collected at a specific location in the indoor space. Due to the discontinuous nature of these signals, they are unable to account for user movement. Therefore, there is a need for techniques to improve the accuracy of indoor positioning in moving situations. This paper proposes a method for enriching all data points with relevant information to obtain improved indoor positioning results and an improved technique for generating movement path data. The proposed technique generates relevant information with continuity by expressing data points as Bounding Boxes (BB) and subsequently clustering them using grid cells. Through the proposed method, we represented and expanded the indoor location data, which consists of data points, as BB. However, data points represented by BB do not have any adjacency information with each other, and as a result, it is not possible to create a movement path between neighboring BB. To solve the problem, we used clustering, a machine learning technique that categorizes data points into groups, to group BBs. This was done by constructing a YOLO (You Only Look Once) grid cell that included all BBs and clustered them into grid cells that were divided into specific regions through the suggestion phase. The clustering results align with the indoor structure of the building, indicating that the data points have been appropriately clustered. BBs inside a grid cell are contained within a moveable area and have related information. We also generated association information between neighboring grid cells through the proposal technique, and then generated path information that can be traveled between grid cells. In this paper, we use the movement path information generated from a dataset as training data for machine learning, and through this, we propose an enhanced indoor positioning technology.

We study the divisibility properties of the constant terms of certain meromorphic modular forms for Hecke groups. We relate those properties to those of some sequences that have already appeared in the literature. For possible use in later drafts, we show how to invert the map taking a Laurent series $f(x) = 1/x + \sum_{n=0}^{\infty}a_{n+1} x^n$ to the sequence of constant terms of its positive powers. At the end of the article, we construct from elementary arithmetic functions some meromorphic but not necessarily modular functions and study their constant terms.

The aim of the study was to examine the long-term prognostic value of changes in cardio-ankle vascular index (CAVI) within a year after coronary artery bypass grafting (CABG). Methods. Patients with coronary artery disease (n = 251) in whom cardio-ankle vascular index (CAVI) was assessed using the VaSera VS-1000 device before and in a year after CABG. Groups with improved CAVI (CAVI decreased from pathological value to normal or the index remained within normal values) or worsened CAVI (persistent pathological index value or an increase in CAVI to pathological values or an increase of 1 unit or more) were identified. The groups were compared for the following events at 10-year follow-up: death from all causes, myocardial infarction, stroke/transient ischemic attack, and the combined endpoint. Results. A year after CABG surgery, 45% of patients showed positive dynamics or stable normal CAVI values. Death from all causes was significantly more common in the group with CAVI progression - 32 (27.6%) than in the group with CAVI improvement - 14 (14.8%; p = 0.029). Patients with CAVI progression were more likely to have MACE (death, MI, stroke/TIA) in 49 (42.2%) cases, compared with patients with CAVI improvement - in 23 (24.5%; p=0.008). Negative dynamics of CAVI (p = 0.024), number of shunts (p = 0.006), and the presence of carotid stenosis more than 30% (p = 0.051) were independent predictors of death from all causes at 10-year follow-up after CABG. The presence of carotid stenosis more than 30% (p = 0.002) and the group with negative dynamics of CAVI after a year (p = 0.008) were independent predictors of the development of the combined endpoint during long-term follow-up. Conclusions. Patients with negative dynamics of CAVI one year after CABG were more likely to experience death from all causes and combined endpoint in the long-term follow-up than patients with normal CAVI. In further studies, it is advisable to evaluate the dynamics of CAVI and, based on this, identify the most effective interventions to improve the prognosis of patients after CABG.

The demand for precise indoor localization services is steadily increasing. Among various methods, fingerprint-based indoor localization has become a popular choice due to its exceptional accuracy, cost-effectiveness, and ease of implementation. However, its performance degrades significantly as a result of multipath signal attenuation and environmental changes. In this paper, we propose an indoor localization method based on fingerprints using self-attention and long short-term memory (LSTM). By integrating a self-attention mechanism and LSTM network, the proposed method exhibits outstanding positioning accuracy and robustness in diverse experimental environments. The performance of the proposed method is evaluated under two different experimental scenarios, which involve 2D and 3D moving trajectories, respectively. The experimental results demonstrate that our approach achieves an average localization error of 1.76 m and 2.83 m in the respective scenarios, outperforming the existing state-of-the-art methods by 42.67% and 31.64%.

The ever-evolving ways attacker continues to im prove their phishing techniques to bypass existing state-of-the-art phishing detection methods pose a mountain of challenges to researchers in both industry and academia research due to the inability of current approaches to detect complex phishing attack. Thus, current anti-phishing methods remain vulnerable to complex phishing because of the increasingly sophistication tactics adopted by attacker coupled with the rate at which new tactics are being developed to evade detection. In this research, we proposed an adaptable framework that combines Deep learning and Randon Forest to read images, synthesize speech from deep-fake videos, and natural language processing at various predictions layered to significantly increase the performance of machine learning models for phishing attack detection.

In industrial settings, gears play a crucial role by assisting various machinery functions such as speed control, torque manipulation, and altering motion direction.The malfunction or failure of these gear components can have serious repercussions, resulting in production halts and financial losses. As a result, there is an increasing requirement to monitor the state of these components in order to avoid such issues from occurring. To address this need, research efforts have focused on early defect detection in gears in order to reduce the impact of possible failures. This study focused on analyzing vibration and thermal datasets from two extruder machine gearboxes using an autoencoder Long Short-Term Memory (LSTM) model. The major goal is to implement an outlier detection approach to detect and classify defects. The results of this study highlighted the extraordinary performance of the Autoencoder LSTM model, which achieved an impressive accuracy rate of 94.42% in recognizing malfunctioning gearboxes within the extruder machine system. Furthermore, the study used a thorough global metrics evaluation methodology to further test the model’s dependability and efficacy, consequently substantiating the proposed approach’s validity.

Deep neural networks (DNNs) have gained prominence in addressing regression problems, offering versatile architectural designs that cater to various applications. In the field of earthquake engineering, seismic response prediction is a critical area of study. Simplified models such as single-degree-of-freedom (SDOF) and multi-degree-of-freedom (MDOF) systems have traditionally provided valuable insights into structural behavior, known for their computational efficiency facilitating faster simulations. However, these models have notable limitations in capturing the nuanced nonlinear behavior of structures and the spatial variability of ground motions. This study focuses on leveraging ambient vibration (AV) measurements of buildings, combined with earthquake (EQ) time-history data, to create a predictive model using a neural network (NN) in image format. The primary objective is to predict a specific building's earthquake response accurately. The training dataset consists of 1,197 MDOF 2D shear models, generating a total of 32,319 training samples. To evaluate the performance of the proposed model, termed MLPER (Machine Learning based Prediction of building structures' Earthquake Response), several metrics are employed. These include mean absolute percentage error (MAPE) and mean deviation angle (MDA) for comparisons in the time domain. Additionally, we assess magnitude-squared coherence values and phase differences (Δφ) for comparisons in the frequency domain. This study underscores the potential of MLPER as a reliable tool for predicting building earthquake response, addressing the limitations of simplified models. By integrating AV measurements and EQ time-history data into a neural network framework, MLPER offers a promising avenue for enhancing our understanding of structural behavior during seismic events, ultimately contributing to improved earthquake resilience in building design and engineering.

Urban rail transit offers advantages like high safety, energy efficiency, and environmental friendliness. With cities rapidly expanding, travelers are increasingly using rail systems, heightening demands for passenger capacity and efficiency while also pressuring these networks. Passenger flow forecasting is an essential part of transportation systems. Short-term passenger flow forecasting for rail transit can estimate future station volumes, providing valuable data to guide operations management and mitigate congestion. This paper investigates short-term forecasting for Suzhou's Shantang Street station. Shantang Street's high commercial presence and distinct weekday versus weekend ridership patterns make it an interesting test case. Wavelet denoising and LSTM modeling were combined to predict short-term flows, comparing the results to standalone LSTM and SVR approaches. This study illustrates that the adopted algorithms exhibit good performance for passenger prediction. The LSTM model with wavelet denoising proved most accurate, demonstrating applicability for short-term rail transit forecasting and practical significance.

The influences of doped hole transporting layer (HTL) on the performances of perovskite solar cells were studied. The influences of electrostatic and surface roughness effect by solvent additive doping were compared. The electrostatic effect of solvent additive doping was compared by addition of 1,8-diiodooctane and 1,8-dichlorooctane to the HTL layer. The surface roughness effect of solvent additive doping was examined using atomic force microscopy. The perovskite crystallinity by doping HTL was studied by X-ray diffraction. The conductivity of HTL was measured using four probe methods and was higher for doped devices. Enhanced conductivity of PEDOT:PSS film was also proved using fluorescence emission quenching. Long-term stability of the device was enhanced by doping HTL. The device with solvent additive exhibited much enhanced stability, retained over 80% of the initial PCE for 400 hours, while PCE of the device without solvent additive decreased to less than 80% after 200 hours. A very weak power microwave treatment was used for post-annealing of the fully fabricated devices. Post-annealing enhanced the lifetime of the device. Comparing different size of cells, the device lifetime was decreased as the cell area increased and as the size of the cell increased, the extent of lifetime enhancement by doping was increased.

Air quality is one of the most concerning problems in major industrialized cities in the world. Prediction of future air quality is highly relevant to public health. In some big cities, multiple air quality measurement stations are deployed at different locations to monitor air pollutants, such as NO2, CO, PM 2.5 and PM 10, over time. At every monitoring time stamp t, we observe one station×feature matrix xt of the pollutant data, which represents a spatio–temporal process. Traditional methods on prediction of air quality typically use data from one station or can only predict a single pollutant (such as PM 2.5) at a time, which ignores the spatial correlation among different stations. Moreover, the air pollution data are typically highly nonstationary, We propose a de-trending graph convolutional LSTM (long short term memory) to continuously predict the whole station×feature matrix in the next 1 to 48 hours, which not only captures the spatial dependency among multiple stations by replacing an inner product with convolution, but also incorporates the de-trending signals (transform a nonstationary process to a stationary one by differencing the data) into our model. Experiments on the air quality data of the city Chengdu and multiple major cities in China demonstrate the feasibility of our method and show promising results.

Background Few data exist on comparative long-term outcomes of severe aortic stenosis (AS) patients with different flow-gradient patterns undergoing transcatheter aortic valve implantation (TAVI). This study sought to evaluate the impact of pre-TAVI flow-gradient pattern on long-term clinical outcomes after TAVI and assess changes in left ventricular ejection fraction (LVEF) of different subtypes of AS patients following TAVI. Methods Consecutive patients with severe AS undergoing TAVI in our institution were screened and prospectively enrolled. Patients were divided in four subgroups according to pre-TAVI flow/gradient pattern: (i) low flow - low gradient (LF-LG): stroke volume indexed (SVi) ≤ 35ml/m2 and mean gradient (MG) < 40mmHg), (ii) normal flow - low gradient (NF-LG): SVi > 35ml/m2 and MG < 40mmHg, (iii) low flow - high gradient (LF-HG) : SVi ≤ 35ml/m2 and MG ≥ 40mmHg and (iv) normal flow - high gradient (NF-HG): SVi > 35 ml/m2 and MG ≥ 40mmHg. Transthoracic echocardiography was repeated at 1-year follow-up. Clinical follow-up was obtained at 12 months, and yearly thereafter until 5-year follow up was complete for all patients. Results A total of 272 patients with complete echocardiographic and clinical follow-up were included in our analysis. Their mean age was 80±7 years and the majority of patients (N=138, 50.2%) were women. 62 patients (22.8% of the study population) were distributed in LF-LG group, 98 patients (36%) were LF-HG patients, 95 patients (34.9%) were NF-HG, while 17 patients (6.3%) were NF-LG. There was a greater prevalence of comorbidities among LF-LG AS patients. One-year all-cause mortality differed significantly between the four subgroups of AS patients (log rank p: 0.022) and was more prevalent among LF-LG patients (25.8%) compared to LF-HG (11.3%), NF-HG (6.3%) and NF-LG patients (18.8%). At 5-year follow-up, global mortality remained persistently higher among LF-LG patients (64.5%) compared to LF-HG (47.9%), NF-HG (42.9%) and NF-LG patients (58.8%) (log rank p: 0.029). At multivariable Cox hazard regression analysis, baseline SVi (HR: 0.951, 95% C.I.; 0.918 – 0.984), the presence of at least moderate tricuspid regurgitation at baseline (HR: 3.091, 95% C.I: 1.645 - 5.809) and at least moderate paravalvular leak (PVL) post-TAVI (HR: 1.456 , 95% C.I.: 1.106 - 1.792 ) were significant independent predictors of late global mortality. LF-LG patients and LF-HG patients exhibited a significant increase in LVEF at 1-year follow-up. A lower LVEF (p<0.001) and a lower Svi (p<0.001) at baseline were associated with LVEF improvement at 1 year. Conclusions Patients with LF-LG AS have acceptable 1-year outcomes with significant improvement in LVEF at 1-year follow-up, but exhibit an exceedingly high 5-year mortality following TAVI. The presence of low transvalvular flow and at least moderate tricuspid regurgitation at baseline and significant paravalvular leak post-TAVI were associated with poorer long-term outcomes in the entire cohort of AS patients. The presence of a low LVEF or a low SVi predicts LVEF improvement at 1-year.

Generally, when cognitive radio users are in Underlay mode, through the analysis of the electromagnetic spectrum in time, frequency, space, and field of multiple dimensional data accounts for judging the inherent correlation of the radio spectrum used/idle state, this is the realization of cognitive radio users (secondary users, SUs) efficient access to the foundation of the limited spectrum resources. Therefore, how to efficiently use of spectrum status of each SU implementation of reception multidimensional combination forecasting is the core of this paper addressing the problem. In this paper, we propose a deep-learning hybrid model called TensorGCN-LSTM based on the tensor data structure. The model treats SUs deployed at different spatial locations under the same frequency and the spectrum status of SUs themselves under different frequencies in the task area as nodes and constructs two types of graph structures. Graph convolutional operations are used to sequentially extract corresponding spatial-domain and frequency-domain features from the two types of graph structures. Then, the long short-term memory (LSTM) model is used to fuse the spatial, frequency, and temporal features of the cognitive radio environment data. Finally, the prediction task of the spectrum distribution situation is accomplished through fully connected layers. Specifically, the model constructs a tensor graph based on the spatial similarity of SUs' locations and the frequency correlation between different frequency signals received by SUs, which describes the electromagnetic wave's dependency relationship in spatial and frequency domains. LSTM is used to capture the electromagnetic wave's dependency relationship in the temporal domain. To evaluate the effectiveness of the model, we conducted ablation experiments on LSTM, GCN, GC-LSTM, and TensorGCN-LSTM models using simulated data. The experimental results show that our model achieved better prediction performance in RMSE, and the correlation coefficient R2 of 0.8753 also confirmed the feasibility of the model.

The deployment of a manufacturing execution system (MES) holds promising potential in facilitating the accumulation of a substantial amount of inspection data. Low quality levels in discrete manufacturing environments are the result of multi-factor coupling and failure to find quality issues in a timely manner within manufacturing settings may trigger the propagation of defects downstream. Currently, most of the inspection quality methods are direct measurement followed by manual judgment. The integration of deep learning methods provides a feasible way to identify defects in a timely manner, thus improving the acceptance rate of factories. This paper focuses on the design of a data-driven quality prediction and control model around discrete manufacturing characteristics, and use fuzzy theory to evaluate the quality level of production stages. Building Multivariate Long and short-term memory learning hidden quality representations to extract predictions from multi-level inspection data in manufacturing systems. Finally, by validating the data of actual produced water dispensers according to three evaluation indexes, RMSE, MAE, MAPE, the results show that Multivariate Long and short-term memory has better prediction performance.

Background and Objective: There is a paucity of literature comparing unilateral instrumented transforaminal lumbar interbody fusion (UITLIF) with bilateral instrumented TLIF (BITLIF) regarding radiological alignment, including the coronal balance, even though UITLIF might have asymmetric characteristics in the coronal plane. This retrospective study aimed to compare the clinical and long-term radiological outcomes of 1-level UITLIF and BITLIF in lumbar degenerative diseases (LDD). Materials and Methods: Patients who underwent 1-level UITLIF with two rectangular polyetheretherketone (PEEK) cages or BITLIF with ≥ 5 years of follow-up at a single hospital were included. We compared the clinical and radiological outcomes between the UITLIF and BITLIF. Results: In total, 63 and 111 patients who underwent UITLIF and BITLIF, respectively, were enrolled. The median follow-up was 85.55 months (range: 60–130). The UITLIF group had a significantly shorter operation time (185.0 [170.0–210.0] vs. 225.0 [200.0–265.0], p &lt; 0.001) and lower estimated blood loss (300.0 [250.0–500.0] vs. 550.0 [400.0–800.0], p &lt; 0.001) than BITLIF group. Regarding the clinical outcomes, there were no significant differences in the intermittent claudication score (p = 0.495) and Kirkaldy–Willis criteria (p = 0.707) at 1 year postoperatively. The interval changes of the local coronal Cobb angle at the index level, L1-S1 lordotic angle, and coronal off-balance from the immediate postoperative radiograph to the last follow-up were not significantly different (p = 0.687, p = 0.701, and p = 0.367, respectively). Conclusion: UITLIF with two rectangular PEEK cages is a viable option for providing comparable clinical outcomes and radiological longevity to BITLIF in 1-level LDD. In addition, UITLIF has advantages over BITLIF in terms of operative time and blood loss.

The performance of bearings plays a pivotal role in determining the dependability and security of rotating machinery. In intricate systems demanding exceptional reliability and safety, the ability to accurately forecast fault occurrences during operation holds profound significance. Such predictions serve as invaluable guides for crafting well-considered reliability strategies and executing maintenance practices aimed at enhancing reliability. In order to ensure the reliability of bearing operation, this article investigates the application of three advanced techniques—Maximum Correlation Kurtosis Deconvolution (MCKD), Multi-Scale Permutation Entropy (MPE), and Long Short-Term Memory (LSTM) recurrent neural networks—for the prediction of the remaining useful life (RUL) of rolling bearings. Each technique's principles, methodologies, and applications are comprehensively reviewed, offering insights into their respective strengths and limitations. Case studies and experimental evaluations are presented to assess their performance in RUL prediction. Findings reveal that MCKD enhances fault signatures, MPE captures complexity, and LSTM excels in modeling temporal patterns. The root mean square error of the prediction results is 0.007. The fusion of these techniques offers a comprehensive approach to RUL prediction, leveraging their unique attributes for more accurate and reliable predictions.

In this paper, we study the constrained minimization problem for an energy functional which is related to the following Kirchhoff type equation \begin{equation*} -\Big(\eta+b\big(\int_{\R^{3}}|\nabla u|^{2}dx\big)^{s}\Big)\Delta u+V(x)u=\mu u +\lambda|u|^{p}u,\end{equation*} where $b$ is a positive constant, parameters $\eta\geq0, \lambda>0$, exponents $s>0$, $0

We propose that the short−term memory (STM), when processing a sentence, uses two uncorrelated processing units in series. The clues for conjecturing this model emerge from studying many novels of the Italian and English Literatures. This simple model, referring to the surface of language, seems to describe mathematically the input-output characteristics of a complex mental process involved in a reading/writing a sentence. We show that there are no significant mathematical/statistical differences between the two literary corpora by considering deep-language variables and linguistic communication channels, therefore, the surface mathematical structure of alphabetical languages is very deeply rooted in human mind, independently of the language used. The first processing unit is linked to the number of words between two contiguous interpunctions, variable Ip, approximately ranging in Miller’s 7±2 range; the second unit is linked to the number of Ip’s contained in a sentence, variable MF, ranging approximately from 1 to 6. The overall capacity required to process fully of a sentence ranges from 8.3 to 61.23 words, values that can be converted into time by assuming a reading speed, giving the range 2.64~19.54 seconds for fast-reading and 5.3~30.1 seconds for average reader. Since a sentence conveys meaning, the surface features we have found might be the starting point to arrive at an Information Theory that includes meaning.

This study investigates the cyclic load application impact on fly ash-based geopolymer composites that are reinforced with a low amount of fibre reinforcement. For reinforcement purposes, PVA and steel fibres are used. For testing purposes, four geopolymer composite mixes were made, 3 of which had fibre reinforcement. Simultaneously specimens were tested for shrinkage, static load-induced creep, and cyclic load-induced creep. For static and cyclic creep testing, specimens were loaded with 20% of their strength. For cyclic creep testing, load application and release cycles were seven days long. When each cycle was introduced, the load was added in steps. In 5 minutes, by 25% steps of the necessary load, the specimens were loaded or unloaded. Only plain specimens show that static creep strains are within cyclic creep strains. For all the other specimens, the static load is higher than the cyclic load-induced creep amplitude. Also, 1% PVA fibre-reinforced specimens show the most elastic characteristics under cyclic load, and 1% steel fibre-reinforced specimens appear to be the most resistant to the cyclic load introduction.

In this paper, four probability functions are compared with the purpose of establishing a methodology to improve the accuracy of the estimated wind energy in a desert city in northwestern Mexico. For the statistical modeling, 3 time series of wind speed data were used, corresponding to the years 2017, 2018, and 2019, recorded with a sonic anemometer at a sampling frequency of 10 Hz. From these, the analysis is performed with different stationarity periods (5, 30, 60, and 600 seconds). The estimation of the parameters characterizing the probability density functions was performed using different methods; the statistical models were evaluated by the coefficient of determination and the Nash-Sutcliffe efficiency coefficient, while their accuracy by the measured quadratic error, the mean square error, the mean absolute error, and the mean absolute percentage error. It was found that Weibull with the energy pattern factor method and Gamma using the method of moments were the probability density functions that best describe the statistical behavior of the wind speed and, therefore, better estimate the energy generated, so it is concluded that the proposed methodology will allow having greater confidence, both in the estimation of the wind speed and in the small wind energy available for its use.

The response of plasmonic metal particles to an electromagnetic wave undergoes significant features at the nanoscale level. Different properties of the internal composition of a metal, such as its ionic background and the free electron gas, begin to manifest more prominently. As the dimensions of the nanostructures decrease, the classical local theory gradually becomes inadequate. Therefore, Maxwellʹs equations need to be supplemented with a relationship determining the dynamics of current density which is the essence of nonlocal plasmonic models. In this field of physics, the standard (linearized) hydrodynamic model (HDM) has been widely adopted with great success, serving as the basis for a variety of simulation methods. However, ongoing efforts are also being made to expand and refine it. Recently, the GNOR (general nonlocal optical response) modification of the HDM has been used, with the intention of incorporating the influence of electron gas diffusion. Clearly, from the classical description of fluid dynamics, a close relationship between viscosive damping and diffusion arises. This offers a relevant motivation for introducing the GNOR modification in an alternative manner. The standard HDM and its existing GNOR modification also do not include the influence of interband electron transitions in the conduction band and other phenomena that are part of many refining modifications of the Drude‐Lorentz and other models of metal permittivity. In this article, we present a modified version of GNOR‐HDM that incorporates the viscosive damping of the electron gas and a generalized Drude‐Lorentz term. In the selected simulations, we also introduce Landau damping which corrects the magnitude of the standard damping constant of the electron gas based on the size of the nanoparticle. We have chosen a spherical particle as a suitable object for testing and comparing HD models and their modifications because it allows finding precise analytical solutions for the interaction and, simultaneously, it is a relatively easily fabricable nanostructure in practice. Our contribution also includes our own analytical method for solving the HDM interaction of a plane wave with a spherical particle. This method forms the core of calculations of the characteristic quantities, such as the extinction cross‐sections and the corresponding components of electric fields and current densities.

This study introduces session-aware recommendation models, leveraging GRU (Gated Recurrent Unit) and attention mechanisms for advanced latent interaction data integration. A primary ad-vancement is enhancing latent context, a critical factor for boosting recommendation accuracy. We address existing models' rigidity by dynamically blending short-term (most recent) and long-term (historical) preferences, moving beyond static period definitions. Our approaches, pre-combination (LCII-Pre) and post-combination (LCII-Post) with fixed (Fix) and flexible learning (LP) weight configurations, are thoroughly evaluated. We conducted extensive experiments to as-sess our models' performance on public datasets such as Amazon and MovieLens 1M. Notably, on the MovieLens 1M dataset, LCII-PreFix achieved a 1.85% and 2.54% higher Recall@20 than II-RNN and BERT4Rec+st+TSA, respectively. On the Steam dataset, LCII-PostLP outperformed these models by 18.66% and 5.5%. Furthermore, on the Amazon dataset, LCII showed a 2.59% and 1.89% improvement in Recall@20 over II-RNN and CAII. These results affirm the significant enhancement our models bring to session-aware recommendation systems, showcasing their po-tential for both academic and practical applications in the field.

This study uses machine learning to predict the convergence results of the Duffing equation with and without damping. The Duffing equation represents a nonlinear second-order differential equation with interesting behavior in undamped free vibration and forced vibration with damping. Convergence alternates randomly between 1 and -1 in undamped free vibration, depending on initial conditions. For forced vibration with damping, multiple factors influence vibration patterns. We utilize the fourth-order Runge-Kutta method to collect convergence results for both conditions. Machine learning techniques, specifically the long short term memory (LSTM) and LSTM-Neural Network (LSTM-NN) method, are employed to predict these convergence values. The LSTM-NN model is a hybrid approach that combines the LSTM method with the addition of hidden layers of neurons. Both the LSTM and LSTM-NN models are thoroughly explored and analyzed in this research. The research process involves three stages: data preprocessing, training, and verification. The results show that the LSTM-NN model becomes more adept at predicting binary datasets, boasting an impressive accuracy of up to 98%. However, when it comes to predicting multiple solutions, the traditional LSTM method outperforms the LSTM-NN approach.

This study explores the application of the Fully Convolutional Network (FCN) algorithm to the field of meteorology, specifically for the short-term nowcasting of severe convective weather events such as hail, convective wind gust (CG) and thunderstorms, short-term heavy rain(STHR) in Gansu. Training data comes from the European Center for Medium-Range Weather Forecasts (ECMWF) and real-time ground observations, the performance of the proposed FCN model based on 2017 to 2021 training datasets demonstrate a high prediction accuracy, with an overall error rate of 16.6%. Furthermore, the model exhibited an error rate of 18.6% across both severe and non-severe weather conditions when testing against the 2022 dataset. Operational deployment in 2023 has yielded an average Critical Success Index (CSI) of 24.3%, a Probability of Detection (POD) of 62.6%, and a False Alarm Ratio (FAR) of 71.2% for these convective events. It is noteworthy that the predicting performance for STHR was particularly effective with the highest POD and CSI, as well as the lowest FAR. CG and hail predictions had comparable CSI and FAR scores, although the POD for CG surpassed that for hail. The FCN model’s optimal performance for hail prediction occurred at the 4th, 8th, and 10th forecast hours, while for CG, the 6th hour was most accurate, and for STHR, the 2nd and 4th hours were most effective. These findings underscore the FCN model's ideal suitability for short-term forecasting of severe convective weather, presenting extensive prospects for the automation of meteorological operations in the future.

The focus of our study was the changes in the composition of semi-natural dry grasslands in Hungary. Since these valuable habitats were created with the help of humans, their preservation also requires regular treatment. Our mowing experiment aimed to investigate the suppression of the native Calamagrostis epigejos L. Roth. In Hungary, in the Cserhát Mountains, eight permanent plots were mowed twice a year. The vegetation was sampled annually between 2001 and 2011. The effects of mowing were investigated with repeated measures analysis of variance (ANOVA). After 10 years, the C. epigejos cover of the mowed plots decreased significantly, from the initial average of 62.38 to 7.50%. Surprisingly, we also experienced a similar decrease in the control plots. Regular treatment also had a positive effect on the number of species and diversity. Species richness increased continuously in both treatment types, which indicates the combined effect of vegetation succession and treatment. The biomass growth of other Poaceae and Fabaceae species, which are important from a grassland management perspective, was also facilitated by mowing. Our results suggest that long-term in situ experiments and continuous botanical studies are needed to establish a reliable utilization of grasslands.

Hundreds of thousands of people living along the Yatsushiro Sea coast have been exposed to methylmercury from the contaminated water of the Chisso factory in Minamata. Methylmercury causes neurological deficits in the cerebrum and cerebellum, with symptoms of ataxia, dysarthria, and visual field constriction in severe cases, but the most common symptom of a milder disorder is somatosensory disturbance due to parietal lobe damage. We investigated the characteristics of the somatosensory disturbances in detail by studying subjective complaints, neurological findings, and quantitative sensory tests in 197 residents (63.2 ± 10.7 years old) from contaminated areas exposed to methylmercury from seafood and 130 residents (63.7 ± 9.3 years old) from control areas. The most commonly affected sensory modalities were superficial peripheral touch and pain in the extremities, followed by two-point discrimination and deep senses, and in the most severe cases, full-body sensory dysfunction and impairment of all sensory submodalities. Although the severity of the sensory submodalities correlated with each other, there was almost no correlation between the severity of these submodalities and indices of peripheral nerve conduction studies, suggesting that the primary cause of the sensory deficits was damage to the central nervous system. The health effects of chronic methylmercury toxicosis can be elucidated by a detailed examination of sensory deficits.

The mathematical structure underlying a sentence is made of two independent processing units in series. The conditional probability density function of sentence length (measured in words) recordable in the extended short–term memory (E–STM) buffer made of CF cells is Gaussian. The number of sentences that authors of Italian and English Literatures have written is compared to that theoretically available to them. The multiplicity factor α describes this comparison. α&gt;1 is more likely than α&lt;1 and often α≫1, therefore the same sentence pattern is reused many times. These texts are easier to read. Very few novels show α&lt;1. In these cases, there are enough diverse sentence patters to convey meaning, but most of them are not used. Texts are more difficult to read, as measured by a universal readability index GU. A mismatch index IM synthetically describes the process. α, GU. and IM increase with year of publication. Future work will concern other literatures – including ancient ones as Greek and Latin – to confirm what, in our opinion, is general feature of human mind.

This study aimed to learn disturbance information to effectively control multiple small unmanned surface vehicles (USVs) for marine research purposes at sea. The learned disturbance information was used to investigate the control of multiple USV swarms based on the leading–following formation technique. Numerous studies have already been conducted on swarm control using drones in the air; however, there has been a paucity of studies on swarm control in the sea. Contrary to swarm control in aviation, the fluid density of the sea is high, and the environment is constantly changing; thus, a robot system would be significantly affected by disturbances. This ultimately influences the maintenance of swarm formations and the accomplishment of missions [1,2]. Therefore, it is difficult to control numerous moving objects while maintaining a formation at sea by using the existing aerial drone swarm control technology. To solve this problem, we learned a long short-term memory (LSTM) model with time-series marine environmental data to predict maritime disturbances. Thereafter, the model was applied to a mission-based position control algorithm of each USV constituting the swarm to design a marine unmanned floating vessel swarm control system that can be applied to various marine environments. Specifically, to effectively control numerous small USVs for marine research purposes, ocean currents are learned using the LSTM algorithm, and the predicted ocean currents are used to generate a swarm USV control system for USV formations. In this study, a control system model of several small USVs equipped with two rear thrusters and a front lateral thruster was designed. The LSTM algorithm was learned using the previous ocean current data, and the velocity of the following ocean currents was predicted. The predictions were then used as a system disturbance to determine the thrust of the controller. To measure ocean currents in the sea when each USV moves, the velocity, azimuth, and position (latitude, longitude) data from the GPS mounted on the USV were used to measure the velocity and direction of the hull's movement. Further, the flow rate was measured using a flow rate sensor on a small USV. The USV movement and position were controlled using an artificial neural network-PID (ANN-PID) controller. This study entailed a comparative analysis of the designed USV model results and those generated by the simulator, including the behavior control rule of the USV swarm and the path of the actual USV swarm at sea. It was verified that the effectiveness of the USV mathematical model and behavior control rules. By comparing the movement path of the swarm USV with and without the disturbance learning algorithm and ANN-PID control algorithm applied to the designed simulator, the position error and maintenance performance of the swarm formation were analyzed, and the application results were compared.

With the escalation of size and complexity of modern Denial of Service attacks, there is a need for research in the context of Machine Learning (ML) used in attack execution and defense against such attacks. This paper investigates the potential use of ML in generating behavioral telemetry data using Long-Short Term Memory network and spoofing requests for the analyzed traffic to look legitimate. For this research, a custom testing environment was built, that listens for mouse and keyboard events and analyzes them accordingly. While the economic feasibility of this attack currently limits its immediate threat, advancements in technology could make it more cost-effective for attackers in the future. Therefore, proactive development of countermeasures remains essential to mitigate potential risks and stay ahead of evolving attack methods.

This work addresses the challenge of EEG visual multiclass classification into 40 classes for Brain-Computer interface applications, by using deep learning architectures. The visual multiclass classification approach offers BCI applications a significant advantage, since it allows the supervision of more than one BCI interaction, considering that each class label supervises a BCI task. However, because of the nonlinearity and nonstationarity of EEG signals, performing multiclassifiers based on EEG features remains a significant challenge for BCI systems. In the present work, mutual information-based discriminant channel selection and Minimum-Norm Estimate algorithms are implemented to select discriminant channels and enhance EEG data. Hence, deep EEGNet and Convolutional-recurrent neural networks are implemented separately to classify EEG data of image visualization into 40 labels. By using the k-fold cross-validation approach, average classification accuracies of 94.8% and 89.8% were obtained by implementing the aforementioned network architectures, respectively. Satisfactory results obtained with this method offer a new implementation opportunity for multi-task BCI applications by utilizing a reduced number of channels (&lt;50%), compared to those presented in the related literature where the whole set of channels is used.

Speech emotion recognition (SER) is a technology that can be applied in distance education to analyze speech patterns and evaluate speakers’ emotional states in real-time. It provides valuable insights and can be used to enhance the learning experience by enabling the assessment of instructors’ emotional stability, a factor that significantly impacts information delivery effectiveness. Students demonstrate different engagement levels during learning activities, and assessing this engagement is an important aspect of controlling the learning process and improving e-learning systems. An important aspect that may influence student engagement is the emotional states of their instructors. Accordingly, this research uses deep learning techniques to create an automated system for recognizing instructors’ emotions in their speech when delivering distance learning. This methodology entails integrating Transformer, convolutional neural network, and long short-term memory architectures into an ensemble to enhance SER. Feature extraction from audio data used Mel-frequency cepstral coefficients, chroma, Mel spectrogram, zero crossing rate, spectral contrast, centroid, bandwidth, roll-off, and root-mean square, with subsequent optimization processes adding noise to, conducting time stretching, and shifting the audio data. Notably, several Transformer blocks were incorporated, and a multi-head self-attention mechanism was employed to identify the relationships between the input sequence segments. The pre-processing and data augmentation methodologies significantly enhanced the precision of the results in that the model achieved accuracy rates of 96.3%, 99.86%, 96.5%, and 85.3% on the Ryerson Audio-Visual Database of Emotional Speech and Song, Berlin Database of Emotional Speech, Surrey Audio-Visual Expressed Emotion, and Interactive Emotional Dyadic Motion Capture datasets. Furthermore, it achieved 83% accuracy on another dataset created for this research—the Saudi Higher Education Instructor Emotions dataset. The results demonstrate this model’s considerable accuracy in detecting emotions in speech data across different languages and datasets.

Accurate streamflow forecasting is crucial for effectively managing water resources, particularly in countries like Colombia, where hydroelectric power generation significantly contributes to the national energy grid. Although highly interpretable, traditional deterministic, physically-driven models often suffer from complexity and require extensive parameterization. Data-driven models like Linear Autoregressive (LAR) and Long Short-Term Memory (LSTM) networks offer simplicity and performance but cannot quantify uncertainty. This work introduces Sparse Variational Gaussian Processes (SVGPs) for forecasting streamflow contributions. The proposed SVGP model reduces computational complexity compared to traditional Gaussian Processes, making it highly scalable for large datasets. The methodology employs optimal hyperparameters and shared inducing points to capture short-term and long-term relationships among reservoirs. Training, validation, and analysis of the proposed approach consider the streamflow dataset from 23 geographically dispersed reservoirs recorded during twelve years in Colombia. Performance assessment reveals that the proposal outperforms baseline Linear Autoregressive (LAR) and Long Short-Term Memory (LSTM) models in three key aspects: adaptability to changing dynamics, provision of informative confidence intervals through Bayesian inference, and enhanced forecasting accuracy. Therefore, the SVGP-based forecasting methodology offers a scalable and interpretable solution for multi-output streamflow forecasting, thereby contributing to more effective water resource management and hydroelectric planning.

To understand dynamics in climate change, informing policy decisions and prompting timely action to mitigate its impact, this study provides a comprehensive analysis of the short-term trend of year-on-year CO2 emission changes across ten countries, considering a broad range of factors including socioeconomic, CO2-related industry, and education. This study uniquely goes beyond the common country-based analysis, offering a broader understanding of the interconnected impact of CO2 emissions across countries. Our preliminary regression analysis, using the ten most significant features, could only explain 66% of variations in the target. To capture emissions trend variation, we categorized countries by the change in CO2 emission volatility (high, moderate, low with upward or downward trends), assessed using standard deviation. We employed machine learning techniques, including feature importance analysis, Partial Dependence Plots (PDPs), sensitivity analysis, and Pearson and Canonical correlation analyses, to identify influential factors driving these short-term changes. The Decision Tree Classifier was the most accurate model, with an accuracy of 96%. It revealed population size, CO2 emissions from coal, the three-year average change in CO2 emissions, GDP, CO2 emissions from oil, education level (incomplete primary), and contribution to temperature rise as the most significant predictors, in order of importance. Furthermore, this study estimates the likelihood of a country transitioning to a higher emission category. Our findings provide valuable insights into the temporal dynamics of factors influencing CO2 emissions changes, contributing to global efforts to address climate change

Recent studies, typically using patient cerebrospinal fluid (CSF), have suggested that different autoantibodies (Aabs) acting on their respective receptors, may underlie neuropsychiatric disorders. The NR1 subunit of the N-methyl-D-aspartate receptor (NMDAR) has been identified as a target of anti-NMDAR Aabs in a number of central nervous system (CNS) diseases including encephalitis and autoimmune epilepsy. However, the role or the nature of Aabs responsible for effects on neuronal excitability and synaptic plasticity is yet to be established fully. Patient-derived anti-NMDAR Aabs have been shown to bind to specific regions within the NR1 subunit; here, peptide immunisation was used to generate Aabs against selected specific NR1 extracellular sequences. ‘Protein A’ purification was used to obtain total IgG and further peptide purification used to obtain a greater percentage of NMDAR-target specific IgG Aabs. Binding and specificity of these anti-NMDAR Aabs was determined using a range of methodologies including enzyme-linked immunosorbent assay, immunocytochemistry and immunoblotting. Functional effects were determined using different in vitro electrophysiology techniques: two-electrode voltage-clamp in Xenopus oocytes and measures of long-term potentiation (LTP) in ex vivo hippocampal brain slices using multi-electrode arrays. We show that anti-NMDAR Aabs generated from peptide immunisation had specificity for NR1 immunisation peptides as well as target-specific binding to the native protein. Anti-NMDAR Aabs had no clear effect on isolated NMDARs in an oocyte expression system. However, we show that peptide-purified anti-NMDAR Aabs prevented the induction of LTP at Schaffer collateral-CA1 synapses in ex vivo brain slices. Our data demonstrate that anti-NMDAR Aabs cause synaptic NMDAR hypofunction at a network level. This work provides a solid basis to address outstanding questions regarding anti-NMDAR Aabs mechanisms of action and, potentially, development of therapies against CNS diseases.

The increasing growth in knowledge about functioning of the nervous system of mammals and humans, as well as the significant neuromorphic technologies development in recent decades, have led to the emergence of a large number of brain-computer interfaces and neuroprosthetics for the regenerative medicine tasks. Neurotechnologies have traditionally been developed for therapeutic purposes to help or replace motor, sensory or cognitive abilities damaged by injury or disease. They also have significant potential for memory enhancement. However, there are still no fully developed neurotechnologies and neural interfaces capable of restoring or expanding cognitive functions, in particular memory, in mammals or humans. In this regard, the search for new technologies in the field of restoration of cognitive functions is an urgent task of modern neurophysiology, neurotechnology and artificial intelligence. Hippocampus is an important brain structure connected to memory and information processing in the brain. The aim of this paper is to propose an approach based on deep neural networks for prediction of hippocampal signals in CA1 region based on received biological input in CA3 region. We compare results of prediction for two widely used deep architectures: reservoir computing (RC) and long short-term memory (LSTM) networks. Proposed study can be viewed as a first step in the complex task of the development of a neurohybrid chip, which allows one to restore memory functions in the damaged rodent hippocampus.

Accurate and reliable mid-long runoff prediction (MLTRP) is of great importance in water re-sources management. However the MLTRP is not suitable in each basin and how to evaluate the applicability of MLTRP is still a question. Therefore, the total mutual information (TMI) index is developed in this study based on the predictor selection method using mutual information (MI) and partial MI (PMI). The relationship between the TMI and the predictive performance of five models is analyzed by applying five models in 222 forecasting scenarios in Australia. The results over 222 forecasting scenarios demonstrate that, compared with the MI, the developed TMI index can better represent the available information in the predictors, and has more significant negative correlation with the RRMSE with the correlation coefficient between -0.62 and -0.85. This means the model's predictive performance will become better along with the increase of TMI, and there-fore the developed TMI index can be used to evaluate the applicability of MLTRP. When TMI is more than 0.1, the available information in the predictors can support the construction of MLTRP models. In addition, the TMI is used to partly explain the difference of predictive performance among five models. In general, the complex models, which can better utilize the contained infor-mation, are more sensitive to the TMI, and have more significant improvement in terms of predic-tive performance along with the increase of TMI. This research can provide support for the study of MLTRP.

This paper proposed emotion recognition methods for consecutive facial images, and implements the inference of neural network model on a field-programmable gate array (FPGA). The proposed emotion recognition methods are based on a neural network model architecture that combines convolution neural networks (CNNs), long short-term memory (LSTM), and fully connected neural networks (FCNNs), called CLDNN or ConvLSTM-FCN. This type of neural network mod-el can analyze the local feature sequences obtained through convolution of data, making it suita-ble for processing time-series data such as consecutive facial images. In this paper, sequences of facial images are sampled from videos corresponding to various emotional state of subjects. The sampled images are then pre-processed with the processes includes facial detection, greyscale conversion, resize, and data augmentation if necessary. The 2-D CNN in ConvLSTM-FCN is used for feature extraction for these pre-processed facial images. These sequences of facial im-age features are time sequences with dependent properties between the elements within them. The LSTM is then used to model these time sequences followed by fully connected neural net-works (FCNNs) for classification. The proposed consecutive facial emotion recognition method achieves an average recognition rate of 99.51% on RAVDESS dataset, 87.80% on BAUM-1s dataset and 96.82% on eNTERFACE’05 data set, using 10-fold cross-validation on the PC. Some com-parisons of recognition accuracies between the proposed method and the other existing related works are conducted in this paper. According to the comparisons, the proposed emotion recog-nition methods outperform the existing related researches. The proposed emotion recognition methods are then implemented on an FPGA chip using the neural network model inference algo-rithms in this paper, and the accuracies of the experiments conducted on the FPGA chip are iden-tical to those obtained on the PC. This verifies that the proposed neural network model imple-mented on the FPGA chip performs well.

A contemporary technique for calculating seismic risk is earthquake nowcasting (EN), which analyzes the progression of the earthquake (EQ) cycle in fault systems. 'Natural time' is a novel idea of time that serves as the foundation for EN assessment. The earthquake potential score (EPS), which has been discovered to have useful uses regionally and worldwide, is a unique tool EN uses to predict seismic risk. For the estimation of the EPS for the occurrences with the highest magnitude among these applications, we have since 2021 concentrated on Greece territory, applying sophisticated Artificial Intelligence (AI) algorithms, both wise (semi)supervised and unsupervised models, along with a customized dynamic sliding window technique that performs as a stochastic filter able to fine-tune the geo seismic occurrences. Long short-term memory (LSTM) neural networks, random forests, and clustering (geospatial) models are three machine learning techniques that are particularly good at finding patterns in vast databases and may be used to enhance earthquake prediction performance. In this study, we attempt to forecast whether practical Machine-learning and AI/Game-theoretic-based approaches can help predict big earthquakes and the normal future seismic cycle for 6-12 months. Specifically, we focus on answering two questions for a given region: (1) Is there a chance that a significant earthquake—say, one with a magnitude of M ≥ 6.0 will happen in the upcoming year? (2) What is the largest earthquake magnitude predicted in the upcoming year, and with which exact geographic coordinates (GCS)? Our results are quite promising and project a high precision accuracy score (≥ 98%) for seismic nowcasting in terms of four predictive parameters: the approximate (a) latitude, (b) longitude, (c) focal depth, and (d) magnitude of the phenomena.