Utilizing Machine Learning Research Articles

Enhancing the Prediction of Flow Characteristics in an Inventive Plate Heat Exchanger Using Deep Learning Techniques

Abstract In modern energy research, minimizing fuel usage and harmful gas emissions are critical priorities. The application of advanced deep learning (DL) models to predict thermohydraulic characteristics of an innovative plate heat exchanger (IPHE) is investigated in this study. Building upon our prior work utilizing machine learning (ML) models, the focus is placed on predicting the Nusselt Number (Nu), friction factor (f), and performance (P) within a Reynolds number range of 500 to 5000. Advanced DL architectures—GRU, LSTM, and CNN—are utilized, resulting in substantial improvements in prediction accuracy and robustness. The LSTM model demonstrates superior performance, achieving R² scores of 0.9986, 0.9985, and 0.9968 for Nu, f, and P, respectively, significantly surpassing prior ML model results of 0.98, 0.979, and 0.9628. The findings highlight the capacity of DL models to capture complex, nonlinear relationships in thermohydraulic data, offering an enhanced approach to optimizing plate heat exchanger (PHE) performance. This work contributes to energy-efficient technological advancements, supporting global efforts to reduce environmental impacts while addressing rising energy demands.

Inhibition of BK channels by GABAb receptors enhances intrinsic excitability of layer 2/3 vasoactive intestinal polypeptide-expressing interneurons in mouse neocortex.

GABAb receptors (GABAbRs) affect many signalling pathways, and hence the net effect of the activity of these receptors depends upon the specific ion channels that they are linked to, leading to different effects on specific neuronal populations. Typically, GABAbRs suppress neuronal activity in the cerebral cortex. Previously, we found that neocortical parvalbumin-expressing cells are strongly inhibited through GABAbRs, whereas somatostatin interneurons are immune to this modulation. Here, we employed in vitro whole-cell patch-clamp recordings to study whether GABAbRs modulate the activity of vasoactive intestinal polypeptide-expressing interneurons (VIP-INs) in layer (L) 2/3 of the mouse primary somatosensory cortex. Utilizing machine learning algorithms (hierarchical clustering and principal component analysis), we revealed that one VIP-IN cluster (about 68% of all VIP-INs) was sensitive to GABAbR activation. Paradoxically, when recordings were performed in standard conditions with high extracellular Ca2+ level, GABAbRs indirectly inhibited the activity of large conductance voltage- and calcium-activated potassium (BK) channels and reduced GABAaR-mediated inhibition, leading to an increase in intrinsic excitability of these interneurons. However, a classical inhibitory effect of GABAbRs on L2/3 VIP-INs was observed in modified artificial cerebrospinal fluid with physiological (low) Ca2+ concentration. Our results are essential for a deeper understanding of mechanisms underlying the modulation of cortical networks. KEY POINTS: Layer 2/3 vasoactive intestinal polypeptide-expressing interneurons (VIP-INs) in the mouse somatosensory cortex cluster into three electrophysiological types differentially sensitive to GABAb receptors (GABAbRs). The majority of VIP-INs (type 1, about 68% of all VIP-INs) are regulated through pre- and postsynaptic GABAbRs, while a subset of these interneurons (types 2 and 3) is controlled only presynaptically. The net effect of GABAbR activation on VIP-IN excitability depends on [Ca2+] in artificial cerebrospinal fluid. When [Ca2+] is high (2.5mM), GABAbRs indirectly inhibit BK channels and reduce GABAaR inhibition leading to increased intrinsic excitability of type 1 VIP-INs. When [Ca2+] is low (1mM), which is more physiological, BK channels do not regulate the intrinsic excitability of VIP-INs and thus postsynaptic GABAbRs canonically decrease the intrinsic excitability of type 1 VIP-INs.

Specific endophenotypes in EEG microstates for methamphetamine use disorder

BackgroundElectroencephalogram (EEG) microstates, which reflect large-scale resting-state networks of the brain, have been proposed as potential endophenotypes for methamphetamine use disorder (MUD). However, current endophenotypes lack refinement at the frequency band level, limiting their precision in identifying key frequency bands associated with MUD.MethodsIn this study, we investigated EEG microstate dynamics across various frequency bands and different tasks, utilizing machine learning to classify MUD and healthy controls.ResultsDuring the resting state, the highest classification accuracy for detecting MUD was 85.5%, achieved using microstate parameters in the alpha band. Among these, the coverage of microstate class A contributed the most, suggesting it as the most promising endophenotype for specifying MUD.DiscussionWe accurately categorize the endophenotype of MUD into different sub-frequency bands, thereby providing reliable biomarkers.

Online Payment Fraud Detection Model Using Machine Learning

The dependability and safety of UPI-based financial transactions are vital for maintaining public confidence, facilitating seamless digital payments, and ensuring economic stability. This paper introduces a software model aimed at identifying and predicting fraudulent behavior in UPI transactions through the use of advanced technologies, which encompass machine learning algorithms, real-time monitoring, and behavioral analysis. The Use of the software is to reduce financial risks by implementing a fraud assessment system that relies on historical transaction data and current information. By utilizing machine learning algorithms along with rule-based methods, the system is able to recognize abnormal transaction patterns, evaluate user behavior, and identify irregularities. The Prediction of model find the risk of transactions based on factors such as the amount, frequency, location, and behavioral patterns of users to highlight potentially suspicious activities. The suggested system functions by implementing four main phases: user engagement, analysis of historical data, continuous monitoring of real-time data, and fraud prediction analysis. This methodology produces comprehensive reports on the risks associated with transactions and offers proactive measures to identify and thwart fraudulent actions within UPI transactions Key Words UPI Transactions, Fraud Detection, Machine Learning, Real-Time Monitoring, Behavioral Analysis, Risk Assessment,

Machine Learning Reveals the Contrasting Roles of Rainfall and Canopy Structure Metrics on the Formation of Canopy Drip and Splash Throughfall

AbstractThroughfall is a significant majority of the total precipitation reaching the ground in forested areas. This study revealed biotic and abiotic factors influencing the throughfall generation process, with the throughfall partitioning into free throughfall, splash throughfall, and canopy drip created at foliar surface drip points (FSDPs) and occasional woody surface drip points (O‐WSDPs), utilizing machine learning. Using a large‐scale rainfall simulator, throughfall drops were simultaneously measured at 19 locations under a mix of deciduous and coniferous tree species in both foliated and unfoliated states. Random forest modeling showed that biotic factors, such as foliage amount, primarily affected the development and volume fraction of canopy drip in foliated trees. In contrast, for unfoliated trees, canopy drip volume fraction was mainly influenced by abiotic factors, including drop size and kinetic energy of open rainfall. The formation and volume fraction of splash throughfall were primarily influenced by abiotic factors for both foliated and unfoliated trees. From the comparison between the foliated and unfoliated states, the generation process of canopy drip was separately clarified between FSDPs and O‐WSDPs. More and larger canopy drip was generated by more foliage with a more wetted canopy with less fluctuation at the FSDPs, whereas a less wetted canopy and/or higher drop impact energy generated more and larger canopy drip at O‐WSDPs. This study underscores the importance of canopy structure and meteorological conditions in determining throughfall partitioning. The findings contribute to a nuanced understanding of rainwater redistribution in forest ecosystems.

On the emergence of machine-learning methods in bottom-up coarse-graining.

Machine-learning methods have gained significant attention in the computational chemistry community as a viable approach to molecular modeling and analysis. Recent successes in utilizing neural networks to learn atomistic force-fields which 'coarse-grain' electronic structure have inspired similar applications to the thermodynamic coarse-graining of chemical and biological systems. In this review, we discuss the current viability and challenges in using machine-learning methods to represent coarse-grained force-fields, as well as the utility of machine-learning in various aspects of coarse-grained modeling.

PENGELOMPOKAN PENDERITA GANGGUAN TIDUR BERDASARKAN GAYA HIDUP MENGGUNAKAN ALGORITMA K-MEANS CLUSTERING

Sleep disorders, including insomnia, can be influenced by various lifestyle factors, such as sleep duration, sleep quality, physical activity, and individual health conditions. This study aims to categorize the risk level of insomnia based on lifestyle using the K-Means clustering algorithm. The data used include sleep duration, sleep quality, heart rate, and daily step count. Through the implementation of the K-Means algorithm, the data is analyzed to group individuals into several categories based on existing lifestyle patterns. The results of the study show a correlation between a healthy lifestyle and better sleep quality. In addition, the resulting clusters provide insight into lifestyle characteristics that affect the risk of insomnia, so that they can be the basis for recommendations for more targeted health interventions. This study is expected to contribute to the development of data-based sleep disorder management strategies by utilizing machine learning methods, especially the K-Means algorithm, to support efforts to improve the quality of life of the community.

Predicting personality traits from Arabic text: an investigation of textual and demographic features with feature selection analysis

Automatic personality recognition (APR) utilizes machine learning to predict personality traits from various data sources. This study aims to predict the big five personality traits from modern standard Arabic (MSA) texts, using both textual and demographic features. The “MSAPersonality” dataset is employed to conduct a comprehensive analysis of features and feature selection methods to evaluate their impact on APR model performance. We compared feature selection algorithms from the filter, wrapper, and embedded-based categories through a systematic experimental design that consisted of feature engineering, feature selection, and regression. This study showed that each trait was more accurately predicted using a distinct set of features. However, age and study level were the most common features among the five traits. Moreover, although there were no statistically significant differences in performance between the feature selection techniques, embedded-based methods offered the best compromise between performance, time, and interpretability. These findings contribute to the understanding of APR in general and among Arabic speakers.

Comparative study of imputation strategies to improve the sarcopenia prediction task.

Sarcopenia, a condition characterized by the progressive loss of skeletal muscle mass and strength, poses significant challenges in research due to missing data. Incomplete datasets undermine the accuracy and reliability of studies, necessitating effective imputation techniques. This study conducts a comparative analysis of three advanced methods-multiple imputation by chained equations (MICE), support vector regression, and K-nearest neighbors (KNN)-to address data completeness issues in sarcopenia research. Following imputation, we utilized machine learning models, including logistic regression, gradient boosting, support vector machine, and random forest, to classify sarcopenia. The methodology encompassed rigorous data preprocessing, normalization, and the synthetic minority oversampling technique to address class imbalance and ensure unbiased model performance. The results revealed substantial variations in model accuracy based on the imputation method employed. The gradient boosting model consistently exhibited superior performance across all imputation strategies, demonstrating its robustness with imputed datasets. Additionally, KNN and MICE emerged as effective imputation techniques, preserving the original data distribution and enabling more accurate classification outcomes. This study underscores the pivotal role of imputation methods in maintaining data integrity and enhancing predictive accuracy in sarcopenia research. The gradient boosting model's reliability across all strategies highlights its potential as a robust classifier, while the suitability of KNN and MICE for preserving data distribution supports their application in similar research contexts. These findings contribute to more reliable and valid insights in sarcopenia studies, ultimately supporting improved clinical outcomes.

A Robust Approach for Pulmonary Disease Diagnosis with Multifractal Features of lung sounds utilizing Machine Learning models

A Robust Approach for Pulmonary Disease Diagnosis with Multifractal Features of lung sounds utilizing Machine Learning models

Enhanced Deep Learning Approaches for Text Classification: A Comprehensive Review

Deep learning models have outperformed traditional machine learning approaches in a variety of text classification tasks, such as sentiment analysis, news categorization, question answering, and natural language inference. Text classification is a widely employed technique in natural language processing, encompassing applications like spam detection, news categorization, information retrieval, sentiment analysis, and intent assessment. Conventional text classifiers utilizing machine learning methods suffer from issues like sparse data, dimensionality challenges, and limited generalization capabilities. In contrast, deep learning-based classifiers address these shortcomings, eliminating the need for intricate feature extraction processes while offering enhanced learning capabilities and increased prediction accuracy. For instance, convolutional neural networks (CNNs). This article outlines the text classification process, with a specific focus on the deep learning models utilized in the context of text classification.

Personalized Cognitive Support via Social Robots

This paper explores the use of personalized cognitive support through social robots to assist the elderly in maintaining cognitive health and emotional well-being. As aging populations grow, the demand for innovative solutions to address issues like loneliness, cognitive decline, and physical limitations increases. The studied social robots utilize machine learning and advanced sensor technology to deliver real-time adaptive interactions, including cognitive exercises, daily task assistance, and emotional support. Through responsive and personalized features, the robot enhances user autonomy and improves quality of life by monitoring physical and emotional states and adapting to the needs of each user. This study also examines the challenges of implementing assistive robots in home and healthcare settings, offering insights into the evolving role of AI-powered social robots in eldercare.

Diagnostics and Prognostics of Boilers in Power Plant Based on Data-Driven and Machine Learning

This paper reports diagnostics and prognostics study of boiler in power plant using actual boiler operating data. This study aims to early detect anomalies that occur in the boiler and to predict the remaining useful life (RUL) after anomalies are detected. The proposed method utilizes machine learning techniques through support vector machine (SVM) and random forest algorithm (RFA) for anomaly detection and similarity-based method of dynamic time warping (DTW) for RUL prediction. The developed method is validated by testing the prediction models using real operating data acquired from three boilers in power plant. The results show that some anomalies are successfully detected by prediction model even though there are anomalies that give low accuracies in predictions. RUL prediction also provides fair results given the limitations of the real data used in building prediction models. Overall, the results of this study have potential to be applied in real system as an auxiliary tool in the boiler condition monitoring to support boiler maintenance programs.

Utilising Machine Learning to Optimise Financial Reporting and Compliance in Sap

Utilising Machine Learning to Optimise Financial Reporting and Compliance in Sap

Machine Learning Guided Lyric-analysis Peer Support Intervention for Psychological Distress in African Population: A BOM Conceptualized Framework

Background Sub-Saharan Africa faces a significant burden of mental health challenges, including depression and anxiety, particularly among vulnerable populations such as women and individuals living with HIV. This study proposes a machine learning-guided Lyric Analysis Peer Support Intervention (LAPSI) to predict psychological distress and inform interventions. The study utilizes machine learning to generate risk profiles, presenting an innovative application in identifying psychological distress risk factors for developing targeted interventions. Objective The objective of this study is to utilize machine learning models to generate risk profiles that predict psychological distress among HIV-positive and HIV-negative populations and to use these profiles to guide thematic development for LAPSI, ensuring confidentiality and informed consent. Methods The study aims to leverage machine learning algorithms such as Logistic Regression and Random Forest to generate risk profiles and to analyze demographic, socio-economic, and psychological factors, including HIV status, age, education, and employment. Introducing a novel application of machine learning in the identification of psychological distress risk factors for intervention development while adhering to ethical standards for handling sensitive data like HIV status. Prior to data collection, ethical approval will be obtained, ensuring participant confidentiality and informed consent. Themes generated from the risk profiles will inform song selection and peer support intervention for LAPSI. Results The machine learning models are expected to highlight education, employment status, and marital status as critical predictors of psychological distress. Geographical and demographic variables, such as district and age, are hypothesized to play a significant role in predicting distress among both HIV-positive and HIV-negative cohorts. Conclusion This study posits that machine learning models will provide actionable insights into predicting psychological distress, enabling targeted interventions through LAPSI. This short communication argues for integrating LAPSI into health policies and calls upon health policy actors to recognize its potential in addressing the mental health crisis in Sub-Saharan Africa, advocating for partnerships with local communities, healthcare providers, and mental health advocates to tailor and implement this intervention effectively.

P-879. Predicting Antimicrobial Resistance and Empiric Treatment Failure in Hospital-Onset Sepsis Using Machine Learning on Electronic Health Records

Abstract Background Antimicrobial resistance (AMR) is a major cause of treatment failure in hospital-onset sepsis. Approaches to guide empiric antimicrobial therapy are urgently needed. We utilized machine learning to predict empiric treatment failure and AMR patterns within a high-risk hospital-onset sepsis cohort using electronic health records (EHRs). Methods We examined hospitalizations with documented bacterial infection from 2010 to 2023 at a tertiary-care academic hospital system. Hospital-onset sepsis was defined using CDC Adult Sepsis Event criteria. Empiric therapy was considered adequate if all pathogens isolated were susceptible to antimicrobials administered. AMR patterns included methicillin (MRSA), extended-spectrum beta-lactamase (ESBL), vancomycin (VRE), ceftriaxone, and carbapenem (CRE). Features for prediction were patient characteristics, hospitalization details, and clinical severity metrics measured within 24 hours prior to sepsis onset. Linear and non-linear machine learning models were evaluated using bootstrap validation and out-of-bag area under the receiver operating characteristic curve (AUC). Results Among 49,581 hospitalizations, 2019 (4%) hospital-onset sepsis encounters were identified. Mean (SD) age was 65 (16.4) years and 900 (45%) were female. At 24 hours after sepsis onset, 377 (19%) received inadequate empiric treatment and 911 (45%) infections expressed AMR. Those who received inadequate empiric treatment were 8.2 (95%CI: 6.2-10.9) times more likely to have an AMR infection. Random forest model was best-performing, with AUC (95%CI) for any AMR pattern, VRE, and CRE at 0.64 (0.63-0.64), 0.71 (0.69-0.72), and 0.70 (0.67-0.73), respectively. The AUC (95%CI) for predicting inadequate empiric antimicrobial treatment was 0.65 (0.64-0.66) with a sensitivity of 64% (58-69%) and specificity of 52% (46-68%) at optimal probability thresholds. Conclusion In a hospital-onset sepsis cohort with prevalent AMR, machine learning approaches utilizing limited EHR data accessible at initial sepsis recognition had moderate, but clinically insufficient, discriminatory ability in predicting AMR and empiric antimicrobial treatment failure. Future research will evaluate the utilization of large language models using unstructured EHRs. Disclosures All Authors: No reported disclosures

Investigation of double geomagnetic storms on 3 and 4 February 2022 using machine learning approach

ABSTRACT The current work utilising machine learning algorithms to investigate the precursor that follows halo coronal mass ejection (CME), eventually leading to moderate geomagnetic storms on the 3rd and 4th of February 2022. The methodology involved developing and testing a machine learning model on collected data, implemented with a Gradient Boosting Regressor (GBR) technique. The GBR algorithm demonstrated strong performance, yielding high accuracy and precision over various error metrics. These findings underscore the potential of machine learning methods to effectively estimate geomagnetic storms. Specifically, they position the GBR algorithm as an optimal choice for this prediction task, outperforming other evaluated options. This study provides evidence for the suitability of the GBR regressor for reliable SYM-H index modelling. These results show that geomagnetic storms may be directly predicted from solar wind parameter data, with a lead time of several days for forecasting, which is important for improving space weather forecasts. According to the study, using the GBR regressor model improves the performance to up to 95%.

Prognostic models of immune-related cell death and stress unveil mechanisms driving macrophage phenotypic evolution in colorectal cancer

BackgroundTumor microenvironment (TME), particularly immune cell infiltration, programmed cell death (PCD) and stress, has increasingly become a focal point in colorectal cancer (CRC) treatment. Uncovering the intricate crosstalk between these factors can enhance our understanding of CRC, guide therapeutic strategies, and improve patient prognosis.MethodsWe constructed an immune-related cell death and stress (ICDS) prognostic model utilizing machine learning methodologies. Furthermore, we performed enrichment analyses and deconvolution algorithms to elucidate the complex interactions between immune cell infiltration and the processes of PCD and stress within a substantial array of transcriptomic data from The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus data base (GEO) related to CRC. Single-cell sequencing and biochemical experiments were used to validate the interaction between the model genes and programmed cell death in tumor cells.ResultsThe ICDS prognostic model exhibited robust predictive performance in seven independent cohorts, revealing an inverse correlation between model scores and patient prognosis. Meanwhile, the ICDS index was positively correlated with clinical stage. Model analysis indicated that patient subgroups with low ICDS index exhibited heightened immune activation features and elevated activity in PCD and stress pathways. Single-cell analysis further revealed that macrophages were the central drivers of immune characteristics underlying prognostic differences within the ICDS prognostic model. Pseudotime analysis and cellular experiments indicated that the model gene GAL3ST4 promotes the transition of macrophages toward an M2 pro-tumor phenotype. Furthermore, cell communication analysis and experimental validation revealed that the cuproptosis in tumor cells suppress GAL3ST4 expression, thereby inhibiting M2-like macrophage polarization.ConclusionIn summary, we constructed the ICDS prognostic model and uncovered the mechanism by which tumor cells downregulate GAL3ST4 expression via cuproptosis to inhibit M2-like macrophage polarization, providing new targets and biomarkers for CRC treatment and prognosis evaluation.

Decoding Pain: A Comprehensive Review of Computational Intelligence Methods in Electroencephalography-Based Brain–Computer Interfaces

Objective pain evaluation is crucial for determining appropriate treatment strategies in clinical settings. Studies have demonstrated the potential of using brain–computer interface (BCI) technology for pain classification and detection. Collating knowledge and insights from prior studies, this review explores the extensive work on pain detection based on electroencephalography (EEG) signals. It presents the findings, methodologies, and advancements reported in 20 peer-reviewed articles that utilize machine learning and deep learning (DL) approaches for EEG-based pain detection. We analyze various ML and DL techniques, support vector machines, random forests, k-nearest neighbors, and convolution neural network recurrent neural networks and transformers, and their effectiveness in decoding pain neural signals. The motivation for combining AI with BCI technology lies in the potential for significant advancements in the real-time responsiveness and adaptability of these systems. We reveal that DL techniques effectively analyze EEG signals and recognize pain-related patterns. Moreover, we discuss advancements and challenges associated with EEG-based pain detection, focusing on BCI applications in clinical settings and functional requirements for effective pain classification systems. By evaluating the current research landscape, we identify gaps and opportunities for future research to provide valuable insights for researchers and practitioners.

Utilizing Machine Learning and causal graph approaches to Address Confounding Factors in Health Science Research: A Scoping Review

Confounding can significantly distort the findings of studies examining cause-and-effect relationships, especially in etiological research. To mitigate this issue, researchers must carefully assess potential confounding variables that may relate to both the exposure and outcome but are not directly influenced by the exposure itself. It is essential that these variables truly impact the outcome rather than simply being correlated with the exposure to avoid false associations. Strengthening confidence in the actual relationship between exposure and outcome requires an understanding of biological mechanisms and the application of various methods to adjust for confounders. The oversight of confounding often arises from inappropriate statistical tests and the aggregation of data across multiple studies. This scoping review article discusses the challenges posed by confounding and presents machine learning approaches for effective control in health science research. Directed acyclic graphs (DAGs) serve as causal graph tools to identify potential confounding variables in health research. By mapping presumed relationships between variables, DAGs enable researchers to estimate causal effects more accurately. While traditional methods such as randomization, matching, and stratification remain effective for controlling confounding, newer techniques like latent variable modeling with negative controls and machine learning methods such as LASSO, Ridge regression, and random forests offer enhanced flexibility and adaptability.