Oscar Lahuerta, Claudio Carretero, Luis Angel Barragan
et al.
This article introduces a hybrid variant of a physics-informed neural network (PINN) that is designed to effectively capture both the rapid dynamics of electrical variables and the slower dynamics of state parameters in a domestic induction heating system. By utilizing observable variables, specifically the voltage and current waveforms from the inductor system, the proposed architecture aims to accurately estimate key electrical parameters, i.e., equivalent resistance and inductance, which vary over time due to the nonlinear magnetic properties of the induction load. To assess the performance of the proposed PINN architecture, a comparison with results obtained using an extended Kalman filter was conducted, which serves as a benchmark for this type of task. In addition, the robustness of both approaches was assessed by introducing varying levels of uncertainty in the observable variables. Finally, the effectiveness of both methods was validated through the analysis of experimental measurements collected from a functional prototype.
Heming Jia, Marjan Kordani, Iman Ahmadianfar
et al.
Precise forecasting of water quality indices (WQI) is essential for safeguarding ecosystems, human health, and sustainable water resource management. This study presents an innovative approach for evaluating river Water Quality Indices using advanced machine learning methods. The approach combines the least squares support vector machine (LSSVM) with the Sherman–Morrison–Woodbury (SMW) formula and local weighting techniques to improve the model's capacity to identify local trends and nonlinearities. The hybrid model, SMW-LSSVM-R, integrates the advantages of SMW-LSSVM with ridge regression to provide a balanced and resilient predictive framework. The model parameters are improved by a self-adaptive teaching-learning-based differential evolution (SATLDE) method, attaining optimal performance. Additionally, SATLDE is combined with a ridge feature selection model to identify the key input factors and boost accuracy. The model also employs optimized multivariate variational mode decomposition (OMVMD) using SATLDE algorithm to more effectively assess complex data patterns. When the models were tested at two Iranian stations, Farisat and Molasani, the SMW-LSSVM-R model with a testing R value of 0.975 and an RMSE of 0.990, exhibited better performance than the basic and OMVMD-enhanced models. These findings demonstrate the potential of the proposed hybrid model to offer valuable insights into environmental monitoring and management.
ABSTRACT Patient outcomes in advanced renal cell carcinoma (RCC) remain poor, with five‐year survival rates ranging from ~10% to 30%. Early projections of therapeutic outcomes could optimize precision medicine and accelerate drug development. While machine learning (ML) models integrating tumor growth inhibition (TGI) metrics have improved survival predictions over traditional models, their application in RCC remains unexplored. Herein, we used TGI metrics and baseline data to evaluate parametric (PM) and semi‐parametric (SPM) survival models alongside ML approaches for predicting progression‐free (PFS) and overall survival (OS) in 1839 RCC patients from four trials (evaluating sunitinib, axitinib, sorafenib, interferon‐alpha, and avelumab + axitinib). Data were split into training (70%) and testing (30%), and feature selection was used to determine parsimonious and robust models. Bootstrap resampling (n = 100) was employed for models' validation, and performance was assessed using C‐index and Integrated Brier Score. In brief, training data results demonstrated that tree‐based ML models (random survival forest (RSF) and XGBoost) outperformed PM and SPM models in predicting PFS (C‐index: 0.783–0.785 vs. 0.725–0.738 for PM and SPM; p < 0.05) and OS (C‐index: 0.77–0.867 vs. 0.750–0.758 for PM and SPM; p < 0.05), with RSF achieving better prediction of PFS and OS using only 3–5 covariates, compared to 9–35 with other tested methods. Tree‐based methods were also superior in the testing data. SHapley Additive exPlanations revealed nonlinear relationships among top predictors, including TGI metrics, underscoring the ability of tree‐based methods to capture complex prognostic interactions. Further validation is required to confirm models' generalizability to additional therapies and patients with differing tumor severity.
Therapeutics. Pharmacology, Public aspects of medicine
The growing adoption of IoT applications has led to increased use of low-power microcontroller units (MCUs) for energy-efficient, local data processing. However, deploying deep neural networks (DNNs) on these constrained devices is challenging due to limitations in memory, computational power, and energy. Traditional methods like cloud-based inference and model compression often incur bandwidth, privacy, and accuracy trade-offs. This paper introduces a novel Decentralized Distributed Sequential Neural Network (DDSNN) designed for low-power MCUs in Tiny Machine Learning (TinyML) applications. Unlike the existing methods that rely on centralized cluster-based approaches, DDSNN partitions a pre-trained LeNet across multiple MCUs, enabling fully decentralized inference in wireless sensor networks (WSNs). We validate DDSNN in a real-world predictive maintenance scenario, where vibration data from an industrial pump is analyzed in real-time. The experimental results demonstrate that DDSNN achieves 99.01% accuracy, explicitly maintaining the accuracy of the non-distributed baseline model and reducing inference latency by approximately 50%, highlighting its significant enhancement over traditional, non-distributed approaches, demonstrating its practical feasibility under realistic operating conditions.
Saskia Denecke, Felix Strakeljahn, Antonia Bott
et al.
Abstract Aetiological models of delusions propose a broad range of predictors. The extent to which these predictors explain variance in persecutory beliefs across the continuum requires systematic investigation. As part of a previous review, 51 aetiological models of delusions were identified in a systematic literature search using PubMed, Web of Science, and Science Direct databases. Omitting repetitions, 66 unique postulated predictors of delusions and persecutory delusions were extracted from these models, of which 55 met our inclusion criteria and were assessed in a cross-sectional online sample stratified by delusion severity (N = 336) using self-report and behavioural measures. Utilising machine learning (i.e., random forests with nested cross-validation), we investigated the extent to which the model-based predictors explain self-reported persecutory beliefs, identified the most relevant predictors, and investigated their specificity in explaining persecutory beliefs as opposed to delusional beliefs or psychopathological symptoms in general. The machine learning model explained 31% of the variance in persecutory beliefs, 47% of delusions in general, and 77% of general psychopathology. The ten predictors with the most influence on predicting persecutory beliefs included negative beliefs about mistrust, cognitive fusion, ostracism, threat anticipation, generalised negative other beliefs, trust, aberrant salience, hallucinations, stress, and emotion regulation difficulties. The limited explanatory power of the proposed predictors raises questions about the validity of existing models and suggests that crucial predictors specific to persecutory delusions may be missing. Our findings highlight the importance of investigating, refining, and cross-validating theoretical aetiological models to improve our understanding of the aetiology of delusions.
M. Graça Pereira, Martim Santos, Renata Magalhães
et al.
University students are at increased risk of developing burnout and psychological distress from high academic workloads and performance expectations. The purpose of this study is to analyze the relationship between psychological and lifestyle variables and academic burnout, as well as to identify burnout risk profiles in psychology students. This study used a cross-sectional design and included 274 Portuguese psychology students, the majority being undergraduates (72.6%). Participants were assessed on psychological well-being, psychological distress, difficulties in emotional regulation, type of diet, physical activity, sleep quality, and burnout. The results showed that psychological distress, difficulties in emotional regulation, and sleep quality were positively associated with burnout, while psychological well-being was negatively associated. Using machine learning algorithms, two distinct profiles were found: “Burnout Risk” and “No Risk”. A total of 62 participants were identified as belonging to the burnout risk profile, showing higher levels of distress, emotional regulation difficulties, poor psychological well-being and sleep quality, pro-inflammatory diet, and less physical activity. The accuracy of the three machine learning models—Random Forest, XGBoost, and Support Vector Machine—was 95.06%, 93.82%, and 97.53%, respectively. These results suggest the importance of health promotion within university settings, together with mental health strategies focused on adaptive psychological functioning, to prevent the risk of burnout.
Machine learning provides more verbose algorithms capable of accurately predicting, classifying groups as needed. Consequently, the objective of this paper is to assess the benchmarking of Supervised Machine Learning Algorithms of K-Nearest Neighbor, Random Forest, Decision Tree and it variants (ID3, C4.5, C5.0 and CART) based on efficiency and performance metrics using python programming after downloading dataset from Kaggle repository. Dataset to the aforementioned models reveals that, the C4.5 variant of decision tree had the highest prediction accuracy, CART and KNN had the minimal learning and prediction time. If accuracy is the based preference, C4.5 variant of decision tree should be recognized, but when the chief concern is nominal time for training and prediction, then CART and KNN standout.
Face mask wearing is a protective health behaviour that helps mitigate the spread of infectious diseases such as influenza and COVID-19. Understanding predictors of face mask wearing can help refine public health messaging and policy in future pandemics. Government mandates influence face mask wearing, but how mandates change predictors of face mask wearing has not been explored. We investigate how mandates changed predictors of face mask wearing and general protective behaviours within Australia during the COVID-19 pandemic using cross-sectional survey data. We compared four machine learning models to predict face mask wearing and general protective behaviours before and after mandates started in Australia; ensemble, tree-based models (XGBoost and random forests) performed best. Other than state, common predictors before and after mandates included age, survey week, average number of contacts, wellbeing, and perception of illness threat. Predictors that only appeared in the top ten before mandates included trust in government, and employment status; and after mandates were willingness to isolate. These distinct predictors are possible targets for future public health messaging at different stages of a new pandemic.
Niloofar Asefi, Leonard Lupin-Jimenez, Tianning Wu
et al.
Reconstructing ocean dynamics from observational data is fundamentally limited by the sparse, irregular, and Lagrangian nature of spatial sampling, particularly in subsurface and remote regions. This sparsity poses significant challenges for forecasting key phenomena such as eddy shedding and rogue waves. Traditional data assimilation methods and deep learning models often struggle to recover mesoscale turbulence under such constraints. We leverage a deep learning framework that combines neural operators with denoising diffusion probabilistic models to reconstruct high-resolution ocean states from extremely sparse Lagrangian observations. By conditioning the generative model on neural operator outputs, the framework accurately captures small-scale, high-wavenumber dynamics even at 99% sparsity (for synthetic data) and 99.9% sparsity (for real satellite observations). We validate our method on benchmark systems, synthetic float observations, and real satellite data, demonstrating robust performance under severe spatial sampling limitations as compared to other deep learning baselines.
Brain-computer interfaces (BCIs) offer transformative potential, but decoding neural signals presents significant challenges. The core premise of this paper is built around demonstrating methods to elucidate the underlying low-dimensional geometric structure present in high-dimensional brainwave data in order to assist in downstream BCI-related neural classification tasks. We demonstrate two pipelines related to electroencephalography (EEG) signal processing: (1) a preliminary pipeline removing noise from individual EEG channels, and (2) a downstream manifold learning pipeline uncovering geometric structure across networks of EEG channels. We conduct preliminary validation using two EEG datasets and situate our demonstration in the context of the BCI-relevant imagined digit decoding problem. Our preliminary pipeline uses an attention-based EEG filtration network to extract clean signal from individual EEG channels. Our primary pipeline uses a fast Fourier transform, a Laplacian eigenmap, a discrete analog of Ricci flow via Ollivier's notion of Ricci curvature, and a graph convolutional network to perform dimensionality reduction on high-dimensional multi-channel EEG data in order to enable regularizable downstream classification. Our system achieves competitive performance with existing signal processing and classification benchmarks; we demonstrate a mean test correlation coefficient of >0.95 at 2 dB on semi-synthetic neural denoising and a downstream EEG-based classification accuracy of 0.97 on distinguishing digit- versus non-digit- thoughts. Results are preliminary and our geometric machine learning pipeline should be validated by more extensive follow-up studies; generalizing these results to larger inter-subject sample sizes, different hardware systems, and broader use cases will be crucial.
High\-cardinality categorical variables pose significant challenges in machine learning, particularly in terms of computational efficiency and model interpretability. Traditional one\-hot encoding often results in high\-dimensional sparse feature spaces, increasing the risk of overfitting and reducing scalability. This paper introduces novel encoding techniques, including means encoding, low\-rank encoding, and multinomial logistic regression encoding, to address these challenges. These methods leverage sufficient representations to generate compact and informative embeddings of categorical data. We conduct rigorous theoretical analyses and empirical validations on diverse datasets, demonstrating significant improvements in model performance and computational efficiency compared to baseline methods. The proposed techniques are particularly effective in domains requiring scalable solutions for large datasets, paving the way for more robust and efficient applications in machine learning.
Abstract This study elucidates the transformative influence of data integration on talent management in the context of evolving technological paradigms, with a specific focus on sustainable practices in human resources. Historically anchored in societal norms and organizational culture, talent management has transitioned from traditional methodologies to harnessing diverse data sources, a shift that enhances sustainable HR strategies. By employing a narrative literature review, the research traces the trajectory of HR data sources, emphasizing the juxtaposition of structured and unstructured data. The digital transformation of HR is explored, not only highlighting the evolution of Human Resource Information Systems (HRIS) but also underscoring their role in promoting sustainable workforce management. The integration of advanced technologies such as machine learning and natural language processing is examined, reflecting on their impact on the efficiency and ecological aspects of HR practices. This paper not only underscores the imperative of balancing data-driven strategies with the quintessential human element of HR but also provides concrete examples demonstrating this balance in action for practitioners and scholars in sustainable human resources.
BackgroundSentiment analysis is a significant yet difficult task in natural language processing. The linguistic peculiarities of Cantonese, including its high similarity with Standard Chinese, its grammatical and lexical uniqueness, and its colloquialism and multilingualism, make it different from other languages and pose additional challenges to sentiment analysis. Recent advances in models such as ChatGPT offer potential viable solutions.
ObjectiveThis study investigated the efficacy of GPT-3.5 and GPT-4 in Cantonese sentiment analysis in the context of web-based counseling and compared their performance with other mainstream methods, including lexicon-based methods and machine learning approaches.
MethodsWe analyzed transcripts from a web-based, text-based counseling service in Hong Kong, including a total of 131 individual counseling sessions and 6169 messages between counselors and help-seekers. First, a codebook was developed for human annotation. A simple prompt (“Is the sentiment of this Cantonese text positive, neutral, or negative? Respond with the sentiment label only.”) was then given to GPT-3.5 and GPT-4 to label each message’s sentiment. GPT-3.5 and GPT-4’s performance was compared with a lexicon-based method and 3 state-of-the-art models, including linear regression, support vector machines, and long short-term memory neural networks.
ResultsOur findings revealed ChatGPT’s remarkable accuracy in sentiment classification, with GPT-3.5 and GPT-4, respectively, achieving 92.1% (5682/6169) and 95.3% (5880/6169) accuracy in identifying positive, neutral, and negative sentiment, thereby outperforming the traditional lexicon-based method, which had an accuracy of 37.2% (2295/6169), and the 3 machine learning models, which had accuracies ranging from 66% (4072/6169) to 70.9% (4374/6169).
ConclusionsAmong many text analysis techniques, ChatGPT demonstrates superior accuracy and emerges as a promising tool for Cantonese sentiment analysis. This study also highlights ChatGPT’s applicability in real-world scenarios, such as monitoring the quality of text-based counseling services and detecting message-level sentiments in vivo. The insights derived from this study pave the way for further exploration into the capabilities of ChatGPT in the context of underresourced languages and specialized domains like psychotherapy and natural language processing.
Computer applications to medicine. Medical informatics, Public aspects of medicine
Mikhail Ivanenko, Damian Wanta, Waldemar T. Smolik
et al.
This study investigated the potential of machine-learning-based stroke image reconstruction in capacitively coupled electrical impedance tomography. The quality of brain images reconstructed using the adversarial neural network (cGAN) was examined. The big data required for supervised network training were generated using a two-dimensional numerical simulation. The phantom of an axial cross-section of the head without and with impact lesions was an average of a three-centimeter-thick layer corresponding to the height of the sensing electrodes. Stroke was modeled using regions with characteristic electrical parameters for tissues with reduced perfusion. The head phantom included skin, skull bone, white matter, gray matter, and cerebrospinal fluid. The coupling capacitance was taken into account in the 16-electrode capacitive sensor model. A dedicated ECTsim toolkit for Matlab was used to solve the forward problem and simulate measurements. A conditional generative adversarial network (cGAN) was trained using a numerically generated dataset containing samples corresponding to healthy patients and patients affected by either hemorrhagic or ischemic stroke. The validation showed that the quality of images obtained using supervised learning and cGAN was promising. It is possible to visually distinguish when the image corresponds to the patient affected by stroke, and changes caused by hemorrhagic stroke are the most visible. The continuation of work towards image reconstruction for measurements of physical phantoms is justified.
By integrating advanced technologies and data-driven systems in smart grids, there has been a significant revolution in the energy distribution sector, bringing a new era of efficiency and sustainability. Nevertheless, with this advancement comes vulnerability, particularly in the form of cyber threats, which have the potential to damage critical infrastructure. False data injection attacks are among the threats to the cyber–physical layer of smart grids. False data injection attacks pose a significant risk, manipulating the data in the control system layer to compromise the grid’s integrity. An early detection and mitigation of such cyberattacks are crucial to ensuring the smart grid operates securely and reliably. In this research paper, we demonstrate different machine learning classification models for detecting false data injection attacks, including the Extra Tree, Random Forest, Extreme Gradient Boosting, Logistic Regression, Decision Tree, and Bagging Classifiers, to secure the integrity of smart grids. A comprehensive dataset of various attack scenarios provides insights to explore and develop effective detection models. Results show that the Extra Tree, Random Forest, and Extreme Gradient Boosting models’ accuracy in detecting the attack outperformed the existing literature, an achieving accuracy of 98%, 97%, and 97%, respectively.
Jorge Buzzio-Garcia, Jaime Vergara, Santiago Rios-Guiral
et al.
In the contemporary cybersecurity landscape, robust attack detection mechanisms are important for organizations. However, the current state of research in Software-Defined Networking (SDN) suffers from a notable lack of recent SDN-OpenFlow-based datasets. This study seeks to bridge this gap by introducing a novel dataset for intrusion detection in Software-Defined Networking named SDNFlow. The dataset, derived from OpenFlow statistics gathered from real traffic, integrates a comprehensive range of network activities. An empirical evaluation leveraging diverse Machine and deep Learning algorithms was performed. Namely, Logistic regression, decision tree, random forest, K-nearest neighbors, Support Vector Machines, and Multilayer Perceptron were tested getting pretty good results with a precision average of 98% to 99% in binary classification and from 97% to 99% in multiclass classification depending of the attack, we highlight the efficacy of K-Nearest Neighbors (KNN) for traffic classification, particularly in detecting DDoS attacks and port scanning. The dataset is valuable for evaluating intrusion detection systems within SDN environments and deepening the understanding of traffic patterns in Software Defined Networks.