Authors - Aneesah Sabar, KA Dilini T Kulawansa Abstract - Federated Learning has emerged as a robust privacy-preserving framework that enables joint model training across multiple distributed clients without sharing raw data. However, the effectiveness of traditional federated learning frameworks is hindered by client heterogeneity, where participants differ in data distribution, computational resources, and communication capabilities. This survey investigates evolution of personalization techniques in Federated Learning that address these challenges by tailoring models to individual clients while maintaining the benefits of global collaboration. The paper categorizes existing personalization approaches into five major groups: local fine-tuning, model interpolation, meta-learning methods, clustered federated learning, and regularization-based techniques. Each method’s core idea, strengths, limitations, and suitability under different heterogeneity conditions are analyzed in detail. The findings indicate that personalization significantly improves fairness, accuracy, and adaptability across heterogeneous clients, though it introduces trade-offs in communication cost, scalability, and privacy. This review concludes that personalization is essential for deploying federated learning in realistic, diverse environments and highlights emerging directions in fairness-aware, resource-efficient, and privacy-preserving personalization. Future research should focus on scalable and dynamic personalization strategies capable of handling evolving client behaviors and large-scale federated systems.
Authors - Bhavanam Sruthi, Mettu Sai Preethi, Krishna Reddy Abstract - Driver drowsiness is a major reason for accidents on the road, hence it is important to detect it early to increase safety on the road. A driver drowsiness detection system based on deep learning algorithms is proposed and it uses images captured through a camera installed inside a car. Various deep learning algorithms, namely CNN, VGG16, DenseNet121, MobileNet, LeNet, AlexNet, RNN, patchTST,Vision Transformer and Swin Transformer are implemented and compared to assess their performance.The system detects the conditions of the driver, whether eyes are open, closed, yawning, or not yawning. Among all these algorithms, the highest accuracy of 97.61% was obtained by using the MobileNet model, which proves that deep learning can play a vital role in detecting drowsiness. In addition, an alert can also be sent to warn the driver.
Authors - Prachita Chaudhari, Shubham Kishor Kadam, Shiwani wagh, Pankajkumar Anawade, Deepak Sharma, Chhitij Raj Abstract - One of such industries is healthcare, where data-driven methods attract much attention, and one among them is the field of predictive analytics that is already making a great difference in the healthcare industry concerning its capacity to enhance early diagnosis and treatment. Through this, whole care is provided, and this implies that the problem of fragmented care addresses systemic issues such as inefficiencies and inflation of costs. Predictive analytics is the most effective in the prediction of risks of disease, i.e. making care a thing related to the individual patient. Besides, it can be utilized to monitor populations and enhance the management of population health utilizing its combination of machine learning, natural language processing and deep learning. Solutions that offer the following benefits, including reduced misdiagnosis, re-source utilization, and affordable access to health care, are also being created with the help of the main enabling technologies including AI, IoT, and big data. Nevertheless, issues like the quality of the data, technical issues, ethical concerns and compliance with laws persist. Future work still to be done areas can be seen in the application of new technologies like quantum computing to answer questions about the public health, real time data that uses IoT, and the application of other mediating technologies in underserved locations that can instill equity and sustainability. Being fueled by the collaboration of various professionals, e.g., clinicians, data scientists, and policy-makers, predictive analytics is bound to enhance patient outcomes and catalyze the better provision of preventive, personalized, and responsibility healthcare solutions.
Authors - Binh Pham Nguyen Thanh, Chau M. Truong, Nhan Thi Cao Abstract - ResNet is widely used in medical image classification due to its strong hierarchical feature extraction capability. This study investigates the integration of Kolmogorov–Arnold Networks (KAN) and ConvKAN into ResNet to analyze the effect of increasing nonlinearity at different stages within a hierarchical skin lesion classification framework. Convolutional KAN is applied at the initial layer to enhance low-level feature extraction, while KAN is introduced at the final layer to improve high-level decision boundary modeling. A combined configuration is also evaluated to examine potential complementary effects across different levels of label granularity. Results show that performance depends on both the integration stage and dataset characteristics. Convolutional KAN at early layers provides limited and inconsistent improvements, whereas KAN at the final layer yields more stable gains. In addition, models incorporating KAN-based architectures generally achieve better performance across metrics such as accuracy, precision, F1-score, and ROC AUC. As classification becomes more fine-grained, Recall consistently decreases despite high ROC AUC, indicating challenges in decision thresholding across hierarchical levels. Overall, KAN is more effective for high-level decision making, while dataset complexity has a greater impact than architectural modifications.
Authors - Mouna Meghana Nagala, Anjan Babu G Abstract - Social Anxiety Disorder (SAD) remains one of the most pervasive mental health challenges globally, characterized by a debilitating “perception gap” where individuals consistently overestimate the visibility of their internal distress while underestimating their social performance. This paper introduces an Explainable AI (XAI) multi-modal sensing system designed for automated social anxiety monitoring and self-perception recalibration. The architecture is founded on an event-driven framework integrating real-time threedimensional facial feature encoding (DeepFace), acoustic prosody extraction (Librosa), and Natural Language Processing (NLP) for cognitive distortion detection. The system implements a Cognitive Behavioral Therapy (CBT) logic layer that provides interpretable feedback on linguistic patterns. System performance was benchmarked against the FER-2013 and RAVDESS repositories, yielding an anxiety detection sensitivity of 92.4% and a specificity of 94.7%. The findings affirm that coupling volumetric affective computing with generative AI constitutes a viable pathway toward trustworthy computer-aided detection (CAD) in behavioral health screening programs.
Authors - Theresa T. Limos, Sheena Sapuay-Guillen Abstract - This study developed PU-Serv: A Tool in Analyzing Student Services Using Machine Learning, a web-based system designed to enhance the evaluation of student services through automated sentiment analysis. The study assessed the existing student services evaluation form in terms of adequacy, efficiency, and reliability and aimed to develop a machine learning–based model to support the analysis of student feedback.A descriptive and developmental research design guided by Agile methodology and the CRISP-DM framework was employed. Data were gathered from focus group discussions, questionnaires, and institutional student feedback records. Natural language processing techniques were used to preprocess narrative feedback, and the Support Vector Machine (SVM) algorithm was integrated into the system due to its high accuracy in sentiment classification. The developed PU-Serv system automatically analyzes student feedback and presents summarized results through a web-based dashboard. The system provides administrators with actionable insights that support data-driven decision-making, helping institutions identify service issues, improve responsiveness, and enhance the overall quality of student services.
Authors - Neel Lathiya, Akshita Kadam, Amit Thakkar Abstract - Industrial tracking tools have led to the development of Quick Response codes, which are an essential component of digital engagement and provide simple access to payments, authentication, and online services with a single scan. However, they are very vulnerable to exploitation, particularly zero-click attacks, which start destructive operations without the user’s consent, due to their architecture, which is based on visual legitimacy, automatic intent execution, and plaintext encoding. This survey looks at the technical aspects of making and reading QR codes, charts the evolution of threats based on QR codes, ranging from physical manipulation to silent deep link hijacking, and explains how these attacks go beyond the robust security models of iOS and Android by utilizing trusted system paths. Based on five significant studies, we analyze real-world attack scenarios, user behavior gaps, and the efficacy of novel defenses like scanner assessment frameworks, zero-trust architecture approaches, and AI-driven payload inspection (AP3X, QRShield). Certain recommendations are made regarding system hardening, cryptographic integration, and user awareness in order to transform QR codes from a latent risk into a safe and verifiable medium.
Authors - Maria Cecilia L. Pangan , Jolou Vincent M. Jala, Ralph Vendel E. Musni, Everly A. Nacalaban, Nenon Roy A. Sandinao, Randy Joy M. Ventayen Abstract - As digital revolution, globalization, and cross-border collaboration re-shape academic landscapes, internationalization of higher education has emerged as a strategic focus for institutions globally. With this, technology-driven innovations particularly learning management systems (LMS), digital platforms, virtual mobility tools and artificial intelligence (AI) have augmented international engagement beyond physical boundaries. Artificial intelligence has immense potential to be a universal technology that boosts product innovation and productivity across a range of industries. This study explores how technological innovations improve internationalization in higher education. Most specifically by how technological innovations improve internationalization in higher education through Technology-Enabled Teaching and Learning, Virtual Mobility and Global Collaboration, Research and Knowledge Exchange, Institutional Governance and Global Competitiveness. Notably, the rapid growth of digital and global academic engagement also causes meaningful implications for students’ and faculty members’ mental health and well-being. This is because when internationalization becomes progressively technology-mediated, matters such as academic pressure, digital fatigue, time-zone differences in global collaboration, and constant online connectivity may contribute to anxiety, stress, and burnout. To attain this goal, the proponents critically examined 156 papers in the body of literature that were indexed by Scopus to examine the advancement of Internationalization in Higher Education through Technology-Driven Innovations Using a systematic review of recent literature, this paper synthesizes global and international perspectives. The findings emphasize that Technology-driven revolutions have redefined the practices and scope of internationalization in higher education. Conversely, obstacles and challenges such as deficient infrastructure, Digital divides, and unfair access to technology deter inclusive involvement and participation
Authors - Gaurav Gupta, Kumar Shashvat, Gunjan Abstract - In India, dengue fever poses a significant threat to public health which continues to worsen. Forecasting methods are crucial to developing effective disease surveillance systems. This study provides an empirical comparison between classical time series forecasting methods, and various machine learning techniques, applied to dengue forecasting for the period of 2013 - 2019 in Chandigarh, India. Seven methods are explored - ARIMA, SARIMA, Exponential Smoothing (ETS), AutoReg, Linear Regression with lagged variables, Decision Tree Regression, and Random Forest Regression. The models are evaluated on multiple criteria which include Mean Absolute Error (MAE), Root Mean Square Error (RMSE), Mean Absolute Percentage Error (MAPE) and Mean Absolute Scaled Error (MASE), and for the statistical models, the Akaike Information Criterion and the Bayesian Information Criterion (AIC/BIC) are used. Random Forest Regression produced the lowest predicted error (MAE 26.95, MASE 0.19), while SARIMA, with seasonal modeling, demonstrated the best and most useful epidemiological interpretability (MAE 45.36, MASE 0.39) of the models. The outcome of the study shows the balance between predictive power of a public health forecasting model, and the interpretability of the model. In this case, SARIMA had the best balance of both and thus, is recommended as the best model for dengue surveillance systems.
Authors - Made Ratih Nurmalasari, Putu Diah Kumalasari, Mirah Candra Adi Saputri Abstract - Through an emphasis on the function of Audit Committee in trying to enhance the correlation of Green Banking Disclosure and also Sustainable Firm Value, this study tried to do investigating how banking firms in Indonesia might have benefits from this practice. As sustainability gets increasingly significant for businesses and also stakeholders alike, banks are under pressure to show transparent of the environmental initiatives. By applying data from Indonesian banks registered in the year of 2021, 2022, and also 2023, this study tended to examine whether banks that actively disclose their efforts of green banking are better allocated to help enhancing their value. The Committee is defined as a moderating variable, shown the significant role to help ensuring good governance and also the disclosures credibility. Data analysis was done by using SPSS with models of multiple regression, as like terms of interaction to help assessing the moderation influences. The outcomes stated that Sustainable Firm Value is greatly improved by Green Banking Disclosure. It is also getting amplified at the time an effective Audit Committee is in place, hoping that good governance is able to increase the influence and also value of sustainability. This research also emphasizes the merging necessity of transparent sustainability measures with strong frameworks of governance to help providing enduring value. The out-comes have actionable information for banking regulators, executives, and also legislators to help integrating sustainability with expansion of corporate.
Authors - Made Ermawan Yoga Antara Abstract - This study is to examine how sustainable MSMEs are impacted by digital innovation and entrepreneurial mindset, mediated by entrepreneurial resilience. This study was carried out in the province of Bali using a quantitative methodology, focusing on MSMEs in the creative economy craft sub-sector. The study sample consisted of 361 MSME owners and leaders selected using proportional random sampling from a total population of 3,745 business units. Data were collected using a Likert-scale questionnaire and analyzed using SEM-PLS with the assistance of SmartPLS software. The results showed that digital innovation and entrepreneurial mindset have a positive and significant effect on both entrepreneurial resilience and sustainability in MSMEs. Additionally, sustainable MSMEs benefit greatly from entrepreneurial resilience. The association between digital innovation and entrepreneurial mindset on sustainable MSMEs is partially mediated by entrepreneurial resilience, according to the mediation test results. Digital innovation has the largest influence on entrepreneurial resilience, while entrepreneurial mindset has the largest direct influence on sustainable MSMEs. These findings emphasize the importance of integrating digital technology adoption and internal entrepreneurial capabilities in driving business sustainability. This research supports dynamic capabilities theory, which emphasizes sensing, seizing, and transforming capabilities in enhancing the resilience and sustainability of MSMEs. Practically, MSMEs need to strengthen digital innovation, entrepreneurial mindsets, and business resilience to adapt to environmental dynamics. This research contributes to the development of sustainable entrepreneurship literature, particularly in the creative economy in developing countries.
Authors - Criscel Jay F. Nayve, Lord Francis B. Navarro,Karen Aparicio Doblas, Elvan Budiongan, Darrel A. Cardana, Max Angelo D. Perin Abstract - This study evaluates online tourist feedback on the Chocolate Hills in Carmen, Bohol, Philippines, using Natural Language Processing (NLP) techniques. Although the destination consistently receives high ratings, negative reviews contain critical insights that can guide tourism management. A total of 4,059 Google Maps reviews were collected, of which 2,011 contained textual content suitable for analysis. The dataset underwent preprocessing using Python and Orange Data Mining before applying sentiment analysis and Latent Dirichlet Allocation (LDA) topic modeling. Results show that, while overall sentiment toward the Chocolate Hills remains strongly positive, negative reviews highlight key concerns related to accessibility, and crowding. Topic modeling identified five dominant themes: scenic appreciation, environmental ambience, crowd density and photo-taking behavior, physical effort required for climbing viewpoints, and perceived cost–benefit value. Sentiment trends from 2020 to 2025 indicate stable positive perceptions despite pandemic-related fluctuations in review volume. Findings suggest that tourists’ satisfaction is primarily driven by the site’s natural beauty, but logistical challenges require targeted management interventions. The study contributes to localized tourism analytics in the Philippines and demonstrates the usefulness of NLP for extracting actionable insights from large volumes of user-generated content.
Authors - Sarvesha Nakharekar, Seedhi Kundap, Suman Madan Abstract - When cyberattacks become ever more extensive and complicated, the demand for intelligent systems capable of executing cyber threat intelligence, digital forensics, and risk management efficiently has increased. We have focused on the important point where digital forensics and cyber threat intelligence meet through this article. In order to build and evaluate the classification models, a publicly accessible intrusion detection dataset was used. The models are Logistic Regression, Decision Tree, Random Forest, Support Vector Machine, K-Nearest Neighbors, and Multilayer Perceptron .The models were evaluated from the perspective of their probable employment in cyber threat intelligence and forensics, based on their performance indicators such as accuracy, precision, recall, F1- score, and computing efficiency .Through a critical discussion, the article also contains a number of significant problems that have been touched upon: the explainability of the attacks, the existence of adversarial attacks, the data imbalance problem, and the limitations of real time processing. The investigation, however, brings up the possibility of using machine learning based on detection outcomes to improve cyber risk management by threat prioritization and thereby making informed decisions. The document is an essential resource for both researchers and field specialists interested in exploring the use of ML to significantly improve threat forecasting, speed incident handling, and strengthen risk management even in a more and more unfriendly domain of cyberattacks.
Authors - Md Tanzid, Md. Foridul Haque, Md. Ismail Hossain, Mohammad Golam Sarowar Abstract - The rice supply chain has been susceptible to quality deterioration, expiry, and a low level of transparency, which are risks to consumer health and food security. To overcome these challenges, the current research suggests a consortium-based blockchain and IoT-enabled smart contract framework to provide a holistic, traceable, and automated governance model. The framework facilitates a consortium of all key stakeholders in the rice supply chain from farmers to retailers as a blockchain network that is co-controlled and resistant to tampering. At key storage infrastructures, Internet of Things (IoT) sensors are deployed to provide the variable storage conditions (humidity and temperature) in real-time that are important in storing rice. The monitoring variables are sent to smart contracts that generate a two-tiered governance system. Upon data showing that a rice lot reached 90% of its shelf life, the intervening automated process will promulgate notifications. Upon expiration of the rice, monitoring will render a smart contract disables the ability to purchase or distribute in the supply chain. The automated process therefore notifies users of public health risks by preventing the introduction and sale of products deemed unsafe for consumption. The framework ensures the sustainable, validated, and tamper evident functionality for continuous monitoring and rule-based execution of perishable products on a public ledger to facilitate enhanced food governance, to lower food safety risks to consumer health, and to promote consumer trust in the rice supply chain.
Authors - Aksh Modi, Agrim Gairola, Suryansh Shah, Sahil Singh, Malvinder Singh Bali Abstract - The increasing degradation of global coral reef ecosystem heavily needs scalable, automated monitoring solution that are capable of operating in resource constrained underwater ecosystem. Though the ongoing State of the Art approaches, such as Vision Transformer and Efficient Net, achieve high classification accuracy, they heavily suffer from computational latency and power requirement that makes them unsuitable for Autonomous Underwater Vehicles (AUVs) or diver held devices. This paper presents a lightweight, real time detection model using the YOLOv8s-cls architecture, which is optimized for edge deployment. Our model achieves a Top 1 Accuracy of 89.84%, conquering the official NOAA Vision Transformer baseline (85.0%) and recent YOLOv8 benchmark at 88.0% accuracy when tested on NOAA-PIFSC-ESD dataset. Crucially, this performance is achieved with a fraction of the computational overhead, enabling high-frequency inference without reliance on cloud connectivity. These results demonstrate that lightweight Convolutional Neural Networks (CNNs) can outperform complex Transformerbased models in texture-centric underwater tasks, providing a viable pathway for immediate, in-situ bleaching assessment by low-power marine robotics.
Authors - Reepu Abstract - Wearable devices are gaining more importance in the present day as they offer various advantages to the users and thereby enabling in generating better outcomes. Wearable devices gained higher focus mainly in the healthcare sector as they support in tracing the critical signs of the patients on a real time basis, provide better support to them as and when needed. The application of wearable devices enables in addressing different health care concerns in an effective manner, the current context witnessed many advancements in the wearable device’s domain. With the integration with other techniques like AI, Internet of Things and other sophisticated tools it can undergo major shift in the health care domain in protecting the lives of the individuals, move from being reactive mode of providing treatment to the patients to more predictive method, enhance clinical skill and outcomes. The overall value of implementing these tools and technologies support in enhancing the life of the patients and supports the practitioners in making better progress in clinical advancements and other areas.
Authors - Shreya Shukla, Mishti Kukreja, Ruchika Katariya Abstract - Attention deficit hyperactivity disorder (ADHD) is a common neurodevelopmental disorder with a prevalence rate ranging between 5 to 7.2% among children and between 2.5 to 6.7% among adults worldwide. In spite of its high prevalence rate, its diagnosis still relies mostly on clinical ratings that have the tendency to show inter-rater differences and confusions with symptoms of other conditions associated with ADHD. Neuroimaging methods, particularly rs-fMRI and sMRI, offer an innovative approach towards providing objective measures for understanding the neurobiological underpinnings of ADHD. This paper offers a systematic narrative review of deep learning methods for ADHD classification using fMRI/sMRI data from 2012 to 2025, with a specific focus on the recent period from 2021 to 2025 characterized by architectural diversity. We classify the literature into three main streams according to the neural networks adopted: (1) Convolutional Neural Networks (CNNs), which involve 2D CNNs, 3D CNNs, residual CNNs, dense CNNs, attention-based CNNs, and graph-based CNNs; (2) Vision Transformers (ViTs), which encompass conventional ViTs, Swin transformers, self-supervised ViTs, multi-modal ViTs, and brain foundation model ViTs; and (3) hybrid CNN-ViT models, which combine both local and global context representations. This work highlights the problems of heterogeneity among multiple sites, inconsistent evaluations, fairness, efficient inference, and clinical deployment. Note: This review does not follow the guidelines for systematic reviews (PRISMA 2020). It is an organized narrative review. Numerical comparisons between different works should be considered approximate due to variations in training/testing sets and data preprocessing.
Authors - Madhumati Pol, Rutuja Chaudhari, Sai Jadhav, Sri Sai Preethi Munnaluri, Rudrani Sarangdhar Abstract - The necessity of developing adaptive, autonomous, and intelligent security systems has developed significantly over time because of the increased volume and complexity of cyber attacks. Therefore, this research project will present an AI-powered multi agent self-evolution cybersecurity intelligence system. The purpose of the system will be for the real-time identification, classification, and prediction of cyber threats. The system will consist of three working agents: Network Monitoring Agent, System-Metrics Surveillance Agent and Threat Intelligence Agent. These agents will be supported by interpretable machine learning classifiers and light-weight Python-based data collection tools. A universal dataset converter will enable it to operate on all types of cybersecurity datasets, and a self-evolving element will allow it to continually update itself with additional information regarding current threats. Dashboards will be provided through the use of Streamlit in order to provide real-time timelines of attacks, CVE intelligence, anomaly detection, and real-time visualization of threats. Results from experimental testing show that the system can improve the accuracy of its threat detection as time progresses and perform well across various datasets. Overall, this work provides a self-learning, scalable, modular, and dataset-agnostic architecture for use within modern enterprise-level cybersecurity environments.
Authors - Elmar B. Noche, Randy Joy M. Ventayen Abstract - Breast cancer remains one of the leading causes of mortality among women, highlighting the need for reliable survival prediction tools. This study applied big data analytics and Cox Proportional Hazards regression to the METABRIC dataset, which contains clinical, pathological, and genomic records from over 2,000 breast cancer patients. Hadoop HDFS was used for distributed storage, while PySpark supported preprocessing and data transformation. After feature selection, six significant predictors were identified: inferred menopausal state, Nottingham Prognostic Index, oncotree code, type of breast surgery, cohort classification, and tumor size. The findings show that combining Hadoop-based infrastructure with interpretable survival modeling can support patient risk stratification, treatment planning, and precision oncology.
Authors - Phat Ly Tan, My Nguyen Kieu, Phung Nguyen Thi Kim Abstract - This paper presents an automated approach for cardiac arrhythmia detection using ECG signals from the CPSC2018 database. The proposed pipeline includes band-pass filtering, normalization, and segmentation of raw ECG recordings, conversion of ECG segments into 2D grayscale images, and multi-label arrhythmia classification using CNN based on a DenseNet architecture. According to the official CPSC2018 labeling scheme, ECG segments are categorized into multiple clinically relevant rhythm types, including normal sinus rhythm and major arrhythmias such as first-degree atrioventricular block, atrial fibrillation, right bundle branch block, left bundle branch block, ventricular ectopic beat, premature atrial contraction, ST-segment elevation, and ST-segment depression. The DenseNet-based architecture combines an oversampling training strategy to alleviate class imbalance. Experimental results on the CPSC2018 database demonstrate the effectiveness of the proposed image-based ECG classification approach, highlighting its potential to assist clinicians in ECG interpretation and early diagnosis of cardiac disorders.
Authors - Renato E. Salcedo, Elmar B. Noche Abstract - The rapid proliferation of Artificial Intelligence (AI) in education has prompted growing scholarly and policy interest in how higher education institutions across the Asia-Pacific region are systematically incorporating AI into teaching, learning, research, and governance. This paper presents a systematic review of 68 peer-reviewed studies, institutional policy documents, and government reports published between 2018 and 2024, examining the extent to which AI is being institutionalized within Asia-Pacific higher education systems in alignment with Education for Sustainable Development (ESD) principles. Using the PRISMA framework and a content analysis methodology, this review identifies four dominant institutionalization pathways: curriculum integration, research infrastructure development, policy formalization, and faculty capacity-building. Findings indicate significant heterogeneity across countries, with East Asian economies particularly China, Japan, and South Korea exhibiting more advanced levels of AI policy coherence, while Southeast Asian and Pacific Island nations remain in nascent stages of formal AI institutionalization. Critical barriers include data governance deficits, algorithmic inequity risks, underfunded professional development pipelines, and insufficient alignment between national AI strategies and ESD frameworks. The review concludes with a set of evidence-based recommendations for regional policymakers, university administrators, and international development organizations to accelerate equitable, sustainable AI institutionalization across the Asia-Pacific higher education landscape.
Authors - Abstract - This research investigated the potential for improving the reliability of the Central Pangasinan Electric Cooperative (CENPELCO) San Carlos 20MVA Substation distribution system through optimized placement of Automatic Circuit Reclosers (ACRs). The study employed a Steady-State Genetic Algorithm (SSGA) implemented in MATLAB to identify optimal ACR locations that minimized SAIDI, SAIFI, and maximized EENS. The algorithm was validated using the IEEE 34-node test feeder, demonstrating its effectiveness in balancing competing objectives. The validated SSGA was then applied to the CENPELCO system, resulting in significant improvements in SAIDI values across all feeders. The research introduced a novel approach to using EENS to represent the number of connected customers at each node, refining the SAIDI calculation and providing a more accurate measure of the impact of outages on consumers. The optimized ACR placement strategy consistently brought SAIDI values below the NEA standard for on-grid electric cooperatives, indicating a substantial enhancement in the reliability of the CENPELCO distribution system. The study provides valuable insights for CENPELCO and other power system operators seeking to improve system reliability and minimize the impact of power outages.
Authors - Phway Phway Aung, Tin Zar Thaw Abstract - Multi-object tracking (MOT) is widely applied in surveillance, traffic monitoring, and autonomous systems. Most MOT systems are created by combining DeepSORT and YOLO series. The original DeepSORT relies on IoU as-sociation-based matching and a fixed age threshold deletion algorithm which of-ten leads to incorrect associations, premature track removal, and frequent ID switches under occlusion or fast motion. To address these limitations, YOLOv9-Based Multi-Object Tracking System is proposed by using GIoU for more reliable geometric association and the enhanced filtering algorithm that are integrated class validation, motion uncertainty estimation with consign similarity, and a Re-Identification (ReID) memory buffer for reducing ID switching. To analyze the performance of the proposed MOT system we compare two cases: IoU and GIoU on Original DeepSORT and the improved DeepSORT and original DeepSORT based on MOT16 videos’ sequences. Experimental evaluation demonstrates that the proposed YOLOv9-Based Multi-Object Tracking System achieves more sta-ble and accurate tracking performance compared to the original DeepSORT system in two cases.
