Abstract: The rapid increase of mobile users and advancement of widely used applications introduce high network demands for low-latency and reliable mobility management in mobile communication networks. However, the traditional handover approaches are rule-based and rely solely on signal strength thresholds with hysteresis margins, which are prone to ping-pong effects and are unable to adapt to dynamic network conditions. Machine Learning (ML) models have been integrated for handover predictions, but their centralized architecture compromises user data privacy, which conflicts with the General Data Protection Regulation (GDPR). These centralized ML approaches also introduce scalability constraints that limit their effectiveness in dense network deployments. To address these challenges, this work proposes a Federated Learning with Software-Defined Mobile Networking (FL-SDMN) framework, a unified approach that integrates federated privacy-preserving learning with centralized network coordination for intelligent handover optimization in 5G and beyond networks. The framework leverages a lightweight federated ExtraTrees ensemble model with weighted tree-based aggregation to preserve data privacy and SDMN to provide global network coordination. It has a three-layer decision pipeline that transforms handover control from a reactive threshold mechanism into a predictive, standards-aligned optimization process. Evaluation of the framework was done with real-world 5G mobility data in terms of decision latency, unnecessary handover reduction, and scalability across diverse network configurations. The findings indicate that the integration of FL, Extra Trees, and SDMN provides a scalable, privacy-preserving, and deployment-ready solution for intelligent mobility management in 5G and beyond networks.

Abstract: Medical image classification in clinical settings is frequently constrained by High-Dimensional, Low-Sample Size (HDLSS) conditions, where conventional Support Vector Machines (SVMs) are geometrically susceptible to the data piling phenomenon, leading to fragile decision boundaries under feature-space perturbations. This study proposes a hybrid CNN-MLP-DWD framework that integrates multi-architecture CNN feature extraction with Distance-Weighted Discrimination (DWD) to address this statistical instability. Pre-trained ResNet50 and DenseNet121 backbones extract complementary representations, fused into a 3072-dimensional vector via Global Average Pooling and horizontal concatenation. A supervised Multi-Layer Perceptron (MLP) bottleneck then compresses this space into a 32-dimensional latent representation, resolving the computational bottleneck of deploying DWD directly on high-dimensional features. Evaluated across breast ultrasound, breast mammography, and chest X-ray datasets, the proposed framework achieves a 29-fold reduction in training latency over the baseline CNN-DWD, elevates breast ultrasound accuracy from 71.83\% to 83.93\% under HDLSS conditions, and attains a macro-AUC of 99.69\% on the chest X-ray benchmark, surpassing all compared methods. Gaussian noise perturbation tests further confirm that DWD maintains better structural resilience over SVM under out-of-distribution clinical conditions.

Abstract: Background: The widespread success of antiretroviral therapy (ART) has transformed HIV into a chronic condition, shifting clinical attention toward aging-associated comorbidities, including autoimmune and rheumatologic diseases. However, the epidemiology, clinical spectrum, and treatment outcomes of these conditions in older people living with HIV (PLH) remain incompletely characterized. Objective: This scoping review aimed to map contemporary evidence on autoimmune and rheumatologic diseases in aging PLH in the ART era, with emphasis on epidemiology, clinical phenotypes, diagnostic challenges, and therapeutic outcomes. Methods: A systematic search of PubMed, Embase, and Scopus was conducted for studies published from January 1, 2021, onward. Eligible studies included adult PLH with autoimmune or rheumatologic conditions and were screened according to PRISMA-ScR methodology. Case reports, non-human studies, and studies published before 2021 were excluded. Data were extracted narratively from included studies. Results: The search yielded 438 records, of which 134 duplicates were removed. After title, abstract, and full-text screening, 8 studies were included. The evidence identified a heterogeneous spectrum of autoimmune and rheumatologic manifestations in PLH, including systemic lupus erythematosus, reactive arthritis, psoriatic arthritis, rheumatic heart disease, immune thrombocytopenia, renal immune-mediated pathology, and myasthenia gravis. Contemporary data suggest that biologic and targeted small-molecule therapies are generally effective and well tolerated in selected PLH, although opportunistic infections and transient viral load increases have been reported with some agents. Conclusion: Autoimmune and rheumatologic diseases in aging PLH represent an emerging ART-era challenge. Prospective studies and multidisciplinary guidelines are needed to optimize diagnosis and treatment.

Abstract: Bundibugyo virus disease, caused by Bundibugyo virus (Orthoebolavirus bundibugyoense), is a severe human ebolavirus disease with substantial mortality, unresolved reservoir ecology, limited diagnostic implementation, and no licensed vaccines or therapeutics specifically approved for this ebolavirus species. The May 2026 public health emergency in the Democratic Republic of the Congo and Uganda renewed the need for a focused synthesis of Bundibugyo virus-specific diagnostics and medical countermeasures. This review synthesizes peer-reviewed literature, preprints, and official public-health documents on diagnostics, antivirals, therapeutics, vaccines, and post-exposure prophylaxis. Comparative evidence from Ebola virus, Sudan virus, Marburg virus, and pan-filovirus platforms is included only where it clarifies Bundibugyo virus-specific evidence, exposes unsupported extrapolation, or defines preparedness gaps. The 2007–2008 outbreak showed that assays optimized for known filoviruses can miss divergent ebolaviruses; the 2026 outbreak underscored the importance of diagnostic breadth, sequencing-based confirmation, decentralized laboratory capacity, and regional coordination. Clinical and immunological data indicate that Bundibugyo virus cannot be reduced to an Ebola virus-like model. Countermeasure evidence remains largely preclinical: recombinant vesicular stomatitis virus vaccines expressing Bundibugyo virus glycoprotein provide the strongest direct animal protection data, whereas antiviral and antibody-based evidence varies widely and requires careful separation of direct Bundibugyo virus data from platform-based extrapolation.

Abstract: Transimpedance amplifiers (TIAs) are the critical current-to-voltage interface in optical receivers, LiDAR front-ends, biomedical sensors, and unconventional applications such as magnetic-resonance receiver-coil arrays and wide-bandgap ultraviolet detectors, and their CMOS design is governed by a fundamental gain-bandwidth-noise trade-off whose structure is rarely made explicit. This review introduces a unifying framework rooted in three explicit assumptions underlying the classical shunt-feedback TIA limit: a single-pole core amplifier (A1), a resistive feedback element (A2), and the full input capacitance loading the feedback summing node (A3). Relaxing one or more of these is shown to be the common structural thread behind every class of bandwidth or noise enhancement in the recent literature, and all surveyed architectures are organized into a six-tier taxonomy, from Tier 0 designs operating within the classical limit to Tier 5 topologies that bypass all three assumptions simultaneously. This taxonomy is supplemented by an orthogonal configurability axis spanning single- and dual-control reconfigurable, variable-gain, and dynamic-range-extension designs. We further show that stability is not removed by these relaxations but migrates with the tier, from the global phase margin of the classical loop to a local regulating loop, a group-delay-flatness constraint, an input-passivity condition, or a multi-loop interaction, so that each architecture carries a predictable stability locus alongside its noise and bandwidth consequences. The taxonomy is cross-referenced with application domains, with closed-form noise-floor boundary plots parametrized by input capacitance and amplifier gain-bandwidth product, and with the CMOS technology landscape, where we argue that the most advanced node is not universally optimal and that node and topology act as complementary rather than competing levers. A single consistent figure of merit, applied uniformly to a representative set of CMOS realizations from 0.6 μm to 16 nm FinFET, shows no monotonic improvement with publication year or node and is presented as a diagnostic indicator rather than an absolute ranking. The review closes with an outlook on 200 Gb/s /lane links, wide-bandgap sensor integration, and the FinFET-to-gate-all-around device transition.

Abstract: A study was conducted to characterize the performance of a HydroFloat® coarse particle flotation (CPF) cell using rougher tailings samples from an industrial copper mining operation. The work involved measuring internal hydrodynamic variables under a wide range of operating conditions. The effect of different operational and hydrodynamic conditions on the metallurgical performance of the HydroFloat® cell was also evaluated. Gas dispersion measurements, such as bubble size distribution, superficial gas velocity (Jg), superficial area flux (Sb), and residence time distribution (RTD), were recorded, enabling a detailed analysis of the cell’s operation. Results show that copper recovery is strongly influenced by the superficial gas velocity (Jg) and the superficial liquid velocity (Jl). It was observed that the bubble diameter (d32) remained relatively constant at 0.5 mm across all operating conditions, which is well below typical bubble sizes for conventional flotation cells. This suggests that contrary to what may be expected, in this kind of machine, small bubbles are able to float coarse particles. Bubble image inspection suggests that the HydroFloat® cell creates conditions conducive to bubble-particle aggregates, which would explain how small bubbles can float coarse particles. This study contributes to the understanding of CPF and establishes a framework for optimization in copper concentrators.

Abstract: Objectives: Paramedics experience high rates of mental disorder symptoms and fatigue, with implications for provider well-being and patient safety. Less attention has been given to sleep quality, even though insomnia is a key contributor to fatigue and is frequently comorbid with other mental health concerns. Our objective was to estimate the prevalence of insomnia symptoms among paramedics in Ontario and examine its association with fatigue, career stage, and comorbidity with other mental disorder symptoms. Methods: We conducted a cross-sectional survey of paramedics from two Ontario paramedic services during compulsory continuing medical education sessions between September and December 2024. Insomnia symptoms were assessed using the Insomnia Severity Index. Fatigue was assessed using selected items from the Copenhagen Burnout Inventory. Mental disorder symptoms were assessed using the posttraumatic stress disorder checklist (PCL-5), patient health questionnaire (PHQ-9) for major depressive disorder, generalized anxiety disorder scale (GAD-7), and alcohol use disorder identification test (AUIDIT). We used multivariable logistic regression to assess demographic risk factors for insomnia and the independent association between insomnia and other mental disorder symptoms. Results: Of 1,019 eligible paramedics, 995 were included in the analysis, yielding a response rate of 96%. Overall, 303 participants (30%) screened positive for clinically significant insomnia, including 242 (24%) with moderate and 61 (6%) with severe symptoms. An additional 396 participants (40%) reported subclinical symptoms. Mid-career (adjusted odds ratio [aOR] 1.94, 95% Confidence Interval [CI] 1.17-3.21) and senior-career (aOR 2.53, 95% CI 1.35-4.75) were at greater risk for insomnia compared to recruits. Insomnia scores were moderately correlated with feeling tired, physically exhausted, and emotionally exhausted. Participants who screened positive for insomnia had increased adjusted odds of depression (aOR 9.08, 95% CI 6.45-12.78), suicidal ideation (aOR 1.99, 95% CI 1.27-3.13), posttraumatic stress disorder (aOR 5.78, 95% CI 4.16-8.01), and hazardous alcohol use (aOR 1.46, 95% CI 1.04-2.05). Conclusions: Insomnia symptoms were highly prevalent among paramedics and strongly associated with fatigue and mental disorder symptoms, particularly depression and suicidal ideation.

Abstract: Universities play a pivotal role in preparing future leaders capable of addressing complex sustainability challenges. However, the development of sustainable entrepreneurial intentions requires more than technical competence; it involves emotional resilience and higher-order cognitive capacity. This study examines how mental wellbeing and academic burnout affect sustainable entrepreneurship attitudes among university students, with systems thinking acting as a mediating competency. Drawing on a sample of 367 students from three universities in northeastern Mexico, Partial Least Squares Structural Equation Modeling (PLS-SEM) was applied to test a six-hypothesis conceptual model. Results confirm that mental wellbeing positively influences both systems thinking (β = 0.31, p < 0.001) and sustainable entrepreneurial orientation (β = 0.28, p < 0.001), while burnout exerts a detrimental effect on both dimensions. Systems thinking partially mediates these relationships, underscoring its role as a cognitive lever for sustainability-oriented action. The model explains 48% of the variance in sustainable entrepreneurship attitude. These findings reinforce the importance of integrating psychological support and cognitive training into higher education programs that aim to promote sustainable entrepreneurship, with direct implications for SDG 4 (Quality Education) and SDG 8 (Decent Work and Economic Growth).

Abstract: The IAPWS-IF97 formulation is the international standard for water and steam thermodynamic properties, but direct formula evaluation without optimization cannot meet the performance demands of compute-intensive applications requiring millions of property calculations. Existing fast direct formula evaluations achieve only limited acceleration through repeated squaring. Alternative methods such as TTSE and SBTL avoid direct evaluation but sacrifice formula-level accuracy. This paper presents two acceleration techniques that directly optimize the IAPWS-IF97 formula evaluation while maintaining full formula accuracy: profiling-guided loop tiling and shared-power scaling. Profiling-guided loop tiling parti tions polynomial summation into cache-friendly tiles with empirically determined boundaries, enabling more effective SIMD vectorization. Shared-power scaling exploits the mathematical relationship between Gibbs/Helmholtz free energy polynomials and their partial derivatives to compute them simultaneously in a single pass, eliminating redundant power calculations. A Rust-based software package, SEUIF97, implements these methods. Benchmark results demonstrate 5x to 36x speedups over baseline implementations and 14x to 30x speedups over CoolProp IF97, the most widely used open-source implementation, with the largest improvements in Regions 1, 2, and 3.

Abstract: In this study, we aimed to evaluate the protein and amino acid contents in grain and the yield of four cowpea genotypes grown in a greenhouse under drought stress at the pollination stage, considered the most sensitive stage to water stress. The four genotypes were two landraces, EPACE 10 and Paulistinha, and two commercial cultivars, Novaera and BR 17 Gurguéia. Water deficit was imposed for eight days, until leaf water potential (Ψa) was close to -2.0 MPa, and then the plants were rehydrated until maturation. On the last day of stress, EPACE 10 and Paulistinha had higher Ψa values than the other two genotypes. When irrigated, landraces had higher grain protein content (GPC) than the commercial cultivars. Under drought, total GPC was reduced only in EPACE 10, while it remained unchanged in BR 17 Gurguéia. In this condition, EPACE 10 exhibited a notably higher grain amino acid content (GAC), with methionine levels peaking. In addition, SDS-PAGE was successfully used to discriminate the genotypes. Under irrigation, the commercial genotypes (BRS Novaera and BR 17 Gurguéia) yielded more than the two landraces. Still, BRS Novaera had lower methionine and cysteine contents. However, under drought, EPACE 10 and BR 17 Gurguéia had the highest yields. In contrast, EPACE 10 had the best amino acid profile, especially for methionine and cysteine, but with slightly lower lysine content. Therefore, EPACE 10 may serve as a potential genotype for improving food security in low-input agriculture, and BR 17 Gurguéia for irrigated farming. However, more studies are needed, especially in the field, to confirm these results.

Abstract: Temporal super-resolution (SR) aims to reconstruct a high-resolution signal from a low-resolution observation. When hardware limits force low sampling rates, this problem becomes critical for non-stationary signals with abrupt transients and rapid spectral changes. This manuscript reports a reproducible case study using pretrained 1D convolutional models and deterministic evaluation on paired real, synthetic, and mixed ECG-like signals. A compact encoder–linear upsampler–refinement architecture is evaluated at 5× upsampling under four training regimes: synthetic-only, real-only, tuned-real, and mixed. Performance is assessed with Mean Squared Error (MSE), Mean Absolute Error (MAE), Log Spectral Distance (LSD), and spectral correlation (SCORR). Across 12 model–dataset combinations, mixed-domain training yields the most robust cross-domain behavior, outperforming single-domain checkpoints on real and mixed evaluation subsets. These findings support the practical value of training-corpus composition for temporal SR under distribution shift. A focused morphological event analysis further shows that reconstruction error concentrates at abrupt amplitude and frequency boundaries, confirming that these transient regions are the dominant local challenge. An exploratory hybrid wavelet–superlet pilot is also reported; it achieves competitive pointwise error on selected domains but exhibits a substantial spectral-fidelity gap, indicating that frequency-aware inputs alone do not guarantee spectral reconstruction without auxiliary spectral losses.

Abstract: In this article, the authors present a methodological approach for solving the Klein-Gordon-Fock equation by means of the separation of variables method, with emphasis on its formulation in the light-front coordinate system. The justification for the study arises from the fact that, although plane-wave solutions are frequently presented in the literature within the context of relativistic quantum mechanics, the solving process leading to their derivation is not always developed explicitly. Initially, they revisit the Klein-Gordon-Fock equation for a free particle in Minkowski spacetime, showing how the separation between the temporal and spatial parts leads to the general solution and allows the interpretation of the components associated with particle and antiparticle propagation. Next, they rewrite the equation in the light-front coordinate system, adopting α_LF=2, and again apply the separation of variables method to the coordinates x⁺, x⁻, and x^⊥. The results show that the solution obtained in this frame preserves the plane-wave structure and recovers, under suitable choices of the superposition coefficients, the wave function expected from the covariant transformation of the scalar product pμxμ. In this way, they demonstrate the coherence between the direct approach through coordinate transformation and the explicit solution of the differential equation. Moreover, the development presented reinforces the didactic potential of separation of variables in the introductory study of relativistic quantum mechanics and indicates the usefulness of the light-front formalism in future treatments involving external potentials, interaction fields, and applications in quantum field theory.

Abstract: Time series forecasting is commonly formulated as a model-centric and single-pass prediction task. However, real-world forecasting often requires task understanding, data diagnosis, contextual feature acquisition, tool-assisted modeling, reflective verification, and human feedback. In this paper, we demonstrate CastClaw, an interactive agent system for context-aware time series forecasting. CastClaw organizes forecasting as a structured runtime workflow that includes intent understanding, data profiling, iterative prediction, reflective verification, and traceable report generation. Supported by a tool library, an execution environment, and state management, CastClaw can invoke forecasting tools, compare models across different families and forecasting skills, track intermediate states, and incorporate user feedback. Through demonstrations on real-world forecasting scenarios, CastClaw shows how forecasting systems can move beyond static predictors toward interactive, evidence-grounded, and verifiable forecasting. Our code is available at (https://github.com/ustc-time-series/CastClaw). The demonstration video could be found at the link (https://ustc-time-series.github.io/cast-claw/).

Abstract: The Standard Model describes particle phenomena through continuous gauge structures, chiral assignments, color, generations, and Yukawa masses, but it does not derive these labels from a deeper structural principle. This paper proposes a carrier-resolution interpretation in which particle species are not separate primitive objects, but different carrier-readable manifestations of one loop-detectable codimension-two archetype defect. The carrier supplies Lorentzian propagation and globally available \(U(1)\) phase closure, while particle labels arise through holonomy, embedding, closure, and saturation conditions. The framework argues that local \(U(1)\) closure favors a neutral-parent starting point, whose persistent asymmetric resolution is modeled as \[P_0\longrightarrow Z_2\oplus (Z_2\!\rightarrow\!Z_3).\] The \(Z_2\) branch is interpreted as lepton-like after Lorentz embedding, whereas the \(Z_2\!\rightarrow\!Z_3\) branch supports a nested confined \(Z_3\) monodromy interpreted as hadronic structure. Incomplete \(Z_3\) sectors are not carrier-readable as isolated hadrons; the usual \(SU(3)_C\) QCD description is retained as the effective high-energy continuum envelope of temporarily resolved \(Z_3\) sectorality. The paper further gives a conditional interpretation of the three observed generations as leading saturation modes of the dominant \(U(1)/Z_2/Z_3\) closure backbone, while higher \(Z_n\) refinements appear as suppressed response corrections rather than ordinary additional generations. As a concrete test, the neutron--proton magnetic-moment ratio is derived from an ideal \(Z_3\)-complete baseline and a rule-generated interface sequence through \(Z_7\). The successive predictions improve from the \(10^{-4}\) level to the few-ppm level and then to below one ppm of observation, without introducing new particles, new fundamental interactions, or fitted coefficients.

Abstract: A quantitative risk assessment of human loss of control over advanced AI used a Bowtie diagram extended with fault tree and event tree analysis. Six primary threats were identified (recursive self‑improvement, power seeking, deceptive alignment, loss of corrigibility, off‑switch subversion, malicious misuse) and six consequences (systemic infrastructure collapse, economic breakdown, resource shortages, non‑human value lock‑in, human marginalization, global supply cascade failures). Preventive and mitigative barriers were assigned per pathway from expert literature. Input probabilities (threat base rates and barrier failure‑on‑demand values) were sourced from experts and modeled with triangular uncertainty distributions. A 1,000‑iteration Monte Carlo simulation propagated epistemic uncertainty, yielding a median probability of the top event (loss of human control) of 12.8% (90% CI: 11.3%–14.4%), roughly 1 in 8. The distribution is approximately symmetric with slight positive skew, indicating modest tail risk if barrier failures interact. Conditional on the top event, Expected Severity is 1.85 on a 1–10 scale (90% CI: 1.75–1.96), suggesting mitigation is effective in most scenarios. Results align with expert estimates and demonstrate barrier effects; narrow CIs reflect model consistency. Remaining tail risks support precautionary governance, increased alignment research, iterative risk modeling, and investment in international coordination with robust safety measures to reduce the existential risk of AI loss of control.

Abstract: To solve the drawbacks of conventional long-cycle wear tests for miniature standing- wave linear ultrasonic motors, an accelerated equivalent wear model and test system were proposed in this work. After primary screening of multiple friction pair materials, graphite and Al2O3 were adopted to modify epoxy films. The optimal friction pair is composed of 6061 hard anodic oxidation film and ECA105 composite film. The matched pair exhibits excellent driving stability and low wear loss, with fatigue wear as the main wear form. Graphite and Al₂O₃ exert synergistic anti-wear and load-bearing effects via forming a stable transfer film on the friction interface. Experimental results confirm that the accelerated test is equivalent to full-life durability test. The presented method and optimized friction pair can effectively guide the development of high-performance ultrasonic motors.

Abstract: Background/Objectives: Diffuse large B-cell lymphoma (DLBCL) is an aggressive lymphoma and one of the most common hematological neoplasia. Entropy is a statistical measure of randomness that can be used to characterize the texture of an input image and measure tissue complexity. Methods: Image processing and computer vision analysis were performed on a series of 114 diagnostic DLBCL cases and 44 reactive lymphoid tissues stained with hematoxylin & eosin (H&E). Histological entropy was measured to differentiate between reactive lymphoid tissue and DLBCL and predict clinical evolution. Gene expression analysis using the NanoString nCounter PanCancer Immune Profiling Panel was performed in 29 cases. Results: Comparison with reactive lymphoid tissue, DLBCL was characterized by lower entropy (7.3 ± 0.2 vs. 6.8 ± 0.6; P < 0.001, respectively). Within the DLBCL diagnostic category and at patient-level analysis, higher entropy was associated with poor overall survival and death events within the first 2 years (hazard-risk = 2.4, P = 0.004) and lower entropy with a moderate and more favorable outcome (hazard-risk = 0.4, P = 0.004). High entropy was also correlated with ECOG performance status ≥ 2, lower protein expression of apoptosis markers of cPARP and cCASP3, and upregulation and downregulation of specific immuno-oncology genes. Conclusion: The histological evaluation of entropy is useful for both the differential diagnosis of reactive lymphoid tissue and DLBCL and can be used as a predictor factor of DLBCL prognosis.

Abstract: The charged-lepton mass hierarchy remains unexplained in the Standard Model: the Higgs mechanism relates \(m_e\), \(m_\mu\), and \(m_\tau\) to three Yukawa couplings, but does not determine their ratios. This paper applies a minimal effective subset of a previously developed reconstruction framework, including carrier embedding, finite-action phase behavior, and codimension-two defect structure, to the charged-lepton mass problem. Charged leptons are modeled as carrier-embedded defect channels arising from a common neutral parent, and their masses are interpreted as quadratic residual responses. The natural variables are therefore root residuals rather than masses themselves. The main algebraic result is that Koide's relation is equivalent to equality between the democratic parent component and the orthogonal channel-splitting component of the charged-lepton root vector. This reduces the hierarchy to a one-angle problem on the Koide cone. The remaining angle is fixed by an effective weak-closure selector involving the neutron--proton mass difference, the democratic electron projection, the beta-continuum scale, the fine-structure constant as a \(U(1)\) carrier-dressing factor, and a finite positive-end recoil term. No continuous parameter is fitted. Using only \(m_e\), \(m_p\), \(m_n\), and \(\alpha\) as physical boundary inputs, the weak-closure selector gives \[m_\mu^{\rm pred}\simeq 105.6565~{\rm MeV}, \qquad m_\tau^{\rm pred}\simeq 1776.94~{\rm MeV}, \] with relative errors of approximately \(-0.0017\%\) and \(+0.0045\%\), respectively. The Koide-cone reduction is algebraic, while the weak-closure selector is presented as an effective boundary condition requiring future microscopic derivation. The result provides a falsifiable route by which the charged-lepton hierarchy may arise from neutral-parent root balance and weak closure rather than from three independent Yukawa parameters.

Abstract: Our first objective was to quantify the associations between health-related events (HRE) before insemination, the relative increase in estrus intensity (REI) at insemination, and the probability of pregnancy per artificial insemination (P/AI) in organic dairy cows. Quanti-fying these associations may aid on-farm decision-making, such as setting the voluntary waiting period, choice of type of semen, do-not-breed and culling decisions. A second ob-jective was to develop predictive models to estimate P/AI based on readily available data, and present common goodness-of-fit results also used in the machine learning community. All data were collected from a certified organic dairy farm in the western USA from 2019 to 2021. Health-related and reproduction data were obtained through DRMS (Raleigh, NC, USA). Activity data were collected using pedometers (IceRobotics, Stirling, UK) mounted on the rear legs. The REI, defined as walking steps per hour before insemination divided by the cow’s baseline steps per hour, was available for 17,238 inseminations from 4,759 cows. The REI was categorized as ≤200% (6,999 inseminations), >200-400% (4,685), >400-600% (2,929), or >600% (2,625). The HRE were available for 65,684 inseminations from 13,365 cows. The HRE was categorized as mastitis (prior to 9,114 inseminations), metabolic (displaced abomasum, ketosis, milk fever; 1,941), reproductive disease (metritis, endometritis, pyometra, retained fetal membranes; 4,907), lameness (4,058), 2 different diseases (4,022), ≥3 different diseases (813), or as healthy (none of these diseases prior to insemination; 40,829). Combinations (COMBO) between REI categories and 0, 1, or ≥2 HRE were also created for 16,415 inseminations in 4,647 cows. Data were split into training and test sets. The training data were used to fit 3 logistic regression models that included either HRE, or REI, or COMBO. Each of the 3 models also included the covariates of prior 3-mo herd P/AI and days in milk (DIM), and the fixed effects of parity, insemination season, days after the previous insemination or days to 1st insemination. A random effect ac-counted for repeated inseminations within cow. Parameter estimates, odds ratios, and the least-square means of the estimated P/AI of the fixed effects were obtained from the logistic regression models. The models’ estimates were applied to the test datasets, and discrimi-nation and calibration statistics were calculated to judge goodness-of-fit. Unadjusted mean P/AI were 31%, 28% and 28% for the HRE, REI and COMBO training datasets. For the HRE model, estimated P/AI ranged from 20% (≥3 different HRE) to 30% (healthy). The estimated P/AI associated with the 4 REI categories were not different from 27% in the REI model. The estimated P/AI associated with the combinations of HRE and REI in the COMBO model varied from 18% after ≥2 HRE and >200-400% REI, to 30% when insemina-tions were in healthy cows with REI >600%. Inseminations in older cows, in the spring, and outside 18-24 d after the previous insemination were also associated with lower esti-mated P/AI. The area underneath the Receiver Operating Characteristic curve ranged from 0.57 (COMBO) to 0.60 (HRE) for the test data, indicating fair discrimination ability of the models. The Brier score ranged from 0.19 to 0.21, indicating moderate performance of the prediction models. Calibration plots showed that the prediction models produced unbi-ased estimated P/AI. In conclusion, the results showed no conclusive evidence of greater estimated P/AI related to greater REI as a measure of estrus activity. More health-related events were associated with lower estimated P/AI. Combinations of low REI and more HRE were associated with notably decreased estimated P/AI. The logistic regression mod-els produce unbiased estimated P/AI. These predictive models may inform insemination and culling decisions in organic dairy cows. A variety of goodness of fit statistics were calculated to allow comparisons of the current logistic regression analyses with future analyses made by other machines learning techniques.

Abstract: The effectiveness of machine learning and deep learning methods for network anomaly detection depends strongly on the quality and representativeness of the datasets used for training and evaluation. However, many publicly available benchmarks rely on synthetic traffic, outdated attack scenarios, or limited representation of encrypted communications. This work presents a network traffic dataset derived from operational firewall logs collected in a heterogeneous institutional environment dominated by HTTPS/TLS traffic. A structured data-centric pipeline was implemented, including preprocessing, behavioral feature engineering, unsupervised pseudo-labeling through the EFMS-KMeans algorithm, class balancing using SMOTE, and temporal sequence generation for sequential analysis. The resulting dataset contains large-scale flow-level records describing volumetric, behavioral, and temporal traffic characteristics while preserving privacy through anonymization procedures. Technical validation was conducted using statistical analysis, entropy-based measurements, clustering quality metrics, and dimensionality reduction techniques, confirming data consistency, diversity, and class separability. The dataset is publicly available through the Mendeley Data repository together with metadata and documentation supporting anomaly detection research, encrypted traffic analysis, and the evaluation of machine learning and deep learning approaches in realistic cybersecurity environments.