Kasthuri Subaramaniam
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Kasthuri Subaramaniam has a total of 37 co-authors. Here are the top co-authors by number of shared publications:
- Oras Baker
Ravensbourne University London
7 shared publications - Sellappan Palaniappan
Faculty of Computing and Digital Technology, HELP University, Kuala Lumpur, Malaysia
6 shared publications Sellappan Palaniappan
HELP University
5 shared publications- Abdul Samad Shibghatullah
Universiti Tenaga Nasional
3 shared publications - Raenu Kolandaisamy
UCSI University
3 shared publications Ngoc Dung Bui
University of Transport and Communications
2 shared publications
Latest Publications
- Open Access
- 14 October 2026
- Published by Multimedia University in Journal of Informatics and Web Engineering
- Vol. 5 (3) , 400-413
Intelligent Telemedicine Systems: Machine Learning for Remote Diagnosis and Patient Management
Telemedicine has experienced significant evolution in the past 10 years from a communication-driven health service to an intelligent, data-driven environment driven by AI (artificial intelligence), ML (machine learning), IoMT (Internet of Medical Things) and cloud–edge computing. This development has been further advanced through ongoing global health challenges that have seen an ageing population, rising burden of chronic, multi-morbid conditions, healthcare specialist shortage, soaring cost of healthcare and major disruptions like the COVID-19 pandemic. Modern telehealth services are now integrating more machine learning algorithms for remote diagnosis, real-time monitoring the patient, predictive risk stratification and proactive treatment. This paper offers a comprehensive academic overview of intelligent telemedicine systems using machine learning for remote diagnosis and patient management. It synthesises and reviews recent literature on image-based diagnostics, physiological time-series modelling, NLP algorithms used to interpret symptoms, and multimodal data fusion models. An integrated system architecture is proposed and presented to illustrate how IoMT devices, data collection pipelines, preprocessing layers, machine learning inbound engines, and clinician decision tools are interlinked in the digital telemedicine landscape. We evaluated a proof-of-concept machine learning workflow for short-term patient deterioration prediction, relying on synthetic IoMT-style patient datasets to demonstrate real-world feasibility and clinical relevance. Apart from technical performance, the paper also describes the challenges of data quality, interoperability, algorithmic bias, explainability, cybersecurity, regulatory compliance, and integration into clinical workflows. Ethical questions, including trust in intelligent healthcare, responsibility, patients’ autonomy, and fair treatment, are raised. Finally, opportunities for future research are mentioned, including digital twins, federated learning, explainable AI, autonomous IoMT ecosystems, and emotionally intelligent telemedicine platforms to guide researchers, clinicians, systems designers, and policy makers on the evolution of telemedicine in a way that is intelligent, proactive and just in health care.
- 14 October 2026
- Published by Multimedia University in Journal of Informatics and Web Engineering
- Vol. 5 (3) , 400-413
Open AccessIntelligent Telemedicine Systems: Machine Learning for Remote Diagnosis and Patient Management
Telemedicine has experienced significant evolution in the past 10 years from a communication-driven health service to an intelligent, data-driven environment driven by AI (artificial intelligence), ML (machine learning), IoMT (Internet of Medical Things) and cloud–edge computing. This development has been further advanced through ongoing global health challenges that have seen an ageing population, rising burden of chronic, multi-morbid conditions, healthcare specialist shortage, soaring cost of healthcare and major disruptions like the COVID-19 pandemic. Modern telehealth services are now integrating more machine learning algorithms for remote diagnosis, real-time monitoring the patient, predictive risk stratification and proactive treatment. This paper offers a comprehensive academic overview of intelligent telemedicine systems using machine learning for remote diagnosis and patient management. It synthesises and reviews recent literature on image-based diagnostics, physiological time-series modelling, NLP algorithms used to interpret symptoms, and multimodal data fusion models. An integrated system architecture is proposed and presented to illustrate how IoMT devices, data collection pipelines, preprocessing layers, machine learning inbound engines, and clinician decision tools are interlinked in the digital telemedicine landscape. We evaluated a proof-of-concept machine learning workflow for short-term patient deterioration prediction, relying on synthetic IoMT-style patient datasets to demonstrate real-world feasibility and clinical relevance. Apart from technical performance, the paper also describes the challenges of data quality, interoperability, algorithmic bias, explainability, cybersecurity, regulatory compliance, and integration into clinical workflows. Ethical questions, including trust in intelligent healthcare, responsibility, patients’ autonomy, and fair treatment, are raised. Finally, opportunities for future research are mentioned, including digital twins, federated learning, explainable AI, autonomous IoMT ecosystems, and emotionally intelligent telemedicine platforms to guide researchers, clinicians, systems designers, and policy makers on the evolution of telemedicine in a way that is intelligent, proactive and just in health care.
- Open Access
- 14 June 2026
- Published by Multimedia University in Journal of Informatics and Web Engineering
- Vol. 5 (2) , 343-360
Synthetic Data Generation for Healthcare and Wellness: Methods, Applications, and Future Directions
Artificial intelligence in healthcare relies heavily upon the availability of high-quality datasets, but very rigid privacy regimes, institutional silos, ethical issues, and heterogeneous data still serve to limit the availability of real-world clinical data. To combat these limitations, you focus on synthetic data generation as a privacy-preserving and scalable alternative to traditional data generation for healthcare and wellness studies. More relevant is a work on a broad and clear framework of synthesis methods based on statistical modelling, rule-based generation and domain-specific clinical logic, in contrast to recent studies that mainly target sophisticated methods for the deep learning architectures. The paper reviews the major synthetic data generation approaches like statistical distributions, machine learning techniques, Generative Adversarial Networks (GANs), Variational Autoencoders (VAEs), diffusion models, and hybrid methods. Furthermore, the paper demonstrates how physiological constraints, time structures, prevalence modelling, and clinically meaningful associations could be incorporated into synthetic datasets in a range of healthcare fields including clinical risk prediction, wearable analytics, mental health text generation, genomics, epidemiological modeling and medical imaging. To ensure reproducibility and accessibility for both research and practice, practical Python-based examples along with domain-aware probabilistic models are provided. Further, it discusses evaluation approaches for evaluating statistical accuracy, downstream utility, and privacy in the process, and stresses their trade-offs on realism, technical efficiency, and disclosure potential. Future avenues for research are also presented, (e.g., digital twins, multimodal patient simulation, long-term disease progression modelling, differentially private generative systems). This work would be a theoretical basis as well as a practical guide for researchers, clinicians, and educators striving to create safe, transparent, and trustworthy synthetic healthcare data for advance of healthcare artificial intelligence.
- 14 June 2026
- Published by Multimedia University in Journal of Informatics and Web Engineering
- Vol. 5 (2) , 343-360
Open AccessSynthetic Data Generation for Healthcare and Wellness: Methods, Applications, and Future Directions
Artificial intelligence in healthcare relies heavily upon the availability of high-quality datasets, but very rigid privacy regimes, institutional silos, ethical issues, and heterogeneous data still serve to limit the availability of real-world clinical data. To combat these limitations, you focus on synthetic data generation as a privacy-preserving and scalable alternative to traditional data generation for healthcare and wellness studies. More relevant is a work on a broad and clear framework of synthesis methods based on statistical modelling, rule-based generation and domain-specific clinical logic, in contrast to recent studies that mainly target sophisticated methods for the deep learning architectures. The paper reviews the major synthetic data generation approaches like statistical distributions, machine learning techniques, Generative Adversarial Networks (GANs), Variational Autoencoders (VAEs), diffusion models, and hybrid methods. Furthermore, the paper demonstrates how physiological constraints, time structures, prevalence modelling, and clinically meaningful associations could be incorporated into synthetic datasets in a range of healthcare fields including clinical risk prediction, wearable analytics, mental health text generation, genomics, epidemiological modeling and medical imaging. To ensure reproducibility and accessibility for both research and practice, practical Python-based examples along with domain-aware probabilistic models are provided. Further, it discusses evaluation approaches for evaluating statistical accuracy, downstream utility, and privacy in the process, and stresses their trade-offs on realism, technical efficiency, and disclosure potential. Future avenues for research are also presented, (e.g., digital twins, multimodal patient simulation, long-term disease progression modelling, differentially private generative systems). This work would be a theoretical basis as well as a practical guide for researchers, clinicians, and educators striving to create safe, transparent, and trustworthy synthetic healthcare data for advance of healthcare artificial intelligence.
- Open Access
- 14 June 2026
- Published by Multimedia University in Journal of Informatics and Web Engineering
- Vol. 5 (2) , 1-33
Synthetic Data Generation for Healthcare Machine Learning: A Case Study in Vital Signs and Diagnostic Predictions
Healthcare Machine Learning (ML) applications face significant challenges in accessing high-quality training data due to stringent privacy regulations, institutional data silos, and concerns over patient confidentiality. This paper explores synthetic data generation as a viable and privacy-preserving alternative to real patient data for developing ML models in healthcare settings. We present techniques for generating realistic vital signs data including body temperature, blood pressure, heart rate, respiratory rate, and oxygen saturation according to appropriate statistical distributions. In addition, we demonstrate how synthetic datasets generated can be used to train diagnostic prediction models. The generated datasets were applied to multiple diagnostic prediction tasks such as hypertension, fever, Chronic Obstructive Pulmonary Disease, atrial fibrillation, and diabetes mellitus. Experimental results reveal that ML models trained solely on synthetic data achieved comparable predictive performance to those trained on real datasets for conditions with explicit physiological manifestations. In particular, gradient boosting classifiers attained an Area Under the Curve (AUC) of up to 0.89 in predicting hypertension. We also illustrate that augmenting sparse real patient data with artificial samples preserves model accuracy at the expense of decreased reliance on sensitive data. This method has great potential to satisfy healthcare organizations who are interested in creating stable ML applications without compromising on privacy standards like Health Insurance Portability and Accountability Act (HIPAA) and General Data Protection Regulation (GDPR).
- 14 June 2026
- Published by Multimedia University in Journal of Informatics and Web Engineering
- Vol. 5 (2) , 1-33
Open AccessSynthetic Data Generation for Healthcare Machine Learning: A Case Study in Vital Signs and Diagnostic Predictions
Healthcare Machine Learning (ML) applications face significant challenges in accessing high-quality training data due to stringent privacy regulations, institutional data silos, and concerns over patient confidentiality. This paper explores synthetic data generation as a viable and privacy-preserving alternative to real patient data for developing ML models in healthcare settings. We present techniques for generating realistic vital signs data including body temperature, blood pressure, heart rate, respiratory rate, and oxygen saturation according to appropriate statistical distributions. In addition, we demonstrate how synthetic datasets generated can be used to train diagnostic prediction models. The generated datasets were applied to multiple diagnostic prediction tasks such as hypertension, fever, Chronic Obstructive Pulmonary Disease, atrial fibrillation, and diabetes mellitus. Experimental results reveal that ML models trained solely on synthetic data achieved comparable predictive performance to those trained on real datasets for conditions with explicit physiological manifestations. In particular, gradient boosting classifiers attained an Area Under the Curve (AUC) of up to 0.89 in predicting hypertension. We also illustrate that augmenting sparse real patient data with artificial samples preserves model accuracy at the expense of decreased reliance on sensitive data. This method has great potential to satisfy healthcare organizations who are interested in creating stable ML applications without compromising on privacy standards like Health Insurance Portability and Accountability Act (HIPAA) and General Data Protection Regulation (GDPR).
- 22 January 2026
- Published by Springer Nature
- p. 139-152
Detecting and Mitigating Adversarial Examples in Neural Networks: An Enhanced PGD Approach
- 22 January 2026
- Published by Springer Nature
- p. 139-152
Detecting and Mitigating Adversarial Examples in Neural Networks: An Enhanced PGD Approach
- Open Access
- 9 December 2025
- Published by MDPI AG in Preprints.org
Mitigating Data Sparsity and Privacy Risks in Educational Recommender System through Federated Learning
The research investigates secure recommender systems through federated learning on educational platforms because online education platforms face increasing threats to student data privacy. The research creates an innovative system which merges FL technology with collaborative filtering to generate personalised course recommendations while maintaining user data protection on client devices. The system evaluated its performance by analysing data from major platforms including edX and Coursera and Udemy and other platforms through MSE and R-squared and precision and recall and F1-score metrics. The evaluation shows that FL maintains user privacy through data aggregation restrictions but users must accept reduced recommendation quality than what centralised systems offer. The research establishes two essential findings which confirm FL maintains user privacy in secure educational settings and reveals that performance reduction from limited data constitutes a core challenge for distributed systems. The research presents two primary methodological contributions which integrate data preprocessing methods for dealing with missing information and develop a complete evaluation system for federated recommendation platforms. The research results differ from previous studies because they demonstrate how model performance deteriorates when operating under federated system constraints. The research develops educational technology FL application expertise by studying privacy-accuracy tradeoffs and presenting methods to boost federated recommender systems in protected data environments.
- 9 December 2025
- Published by MDPI AG in Preprints.org
Open AccessMitigating Data Sparsity and Privacy Risks in Educational Recommender System through Federated Learning
The research investigates secure recommender systems through federated learning on educational platforms because online education platforms face increasing threats to student data privacy. The research creates an innovative system which merges FL technology with collaborative filtering to generate personalised course recommendations while maintaining user data protection on client devices. The system evaluated its performance by analysing data from major platforms including edX and Coursera and Udemy and other platforms through MSE and R-squared and precision and recall and F1-score metrics. The evaluation shows that FL maintains user privacy through data aggregation restrictions but users must accept reduced recommendation quality than what centralised systems offer. The research establishes two essential findings which confirm FL maintains user privacy in secure educational settings and reveals that performance reduction from limited data constitutes a core challenge for distributed systems. The research presents two primary methodological contributions which integrate data preprocessing methods for dealing with missing information and develop a complete evaluation system for federated recommendation platforms. The research results differ from previous studies because they demonstrate how model performance deteriorates when operating under federated system constraints. The research develops educational technology FL application expertise by studying privacy-accuracy tradeoffs and presenting methods to boost federated recommender systems in protected data environments.
- Open Access
- 14 October 2025
- Published by Multimedia University in Journal of Informatics and Web Engineering
- Vol. 4 (3) , 64-89
Training the Brain: A Machine Learning Approach to Predicting Wellbeing Through Intentional Thought Pattern Modification
This study provides a quantitative framework for wellbeing outcome prediction through intentional cognitive pattern alteration. We demonstrated 81.67% accurate prediction of wellbeing states, in a three-level classification (Low, Medium, High), using a Random Forest classifier with 16 features from psychological, physiological, and behavioural metrics. Our model singles out the gratitude cultivation (21.3%) and peace duration (23.7%) as the strongest predictors of positive well-being outcomes, which provides empirical support to traditional approaches of cognitive training with empirical evidence. Analysis of 1,000 synthetic cases shows that consistent practice of positive thought patterns over 3-6 months can strongly shift wellbeing states, with key behavioural markers showing progressive improvement which include increased joy moments, reduced anxiety episodes, and enhanced sleep quality. Our results establish that cognitive training outcomes can be quantitatively tracked and predicted with meaningful accuracy, hence providing a data-driven approach to mental health intervention design. Additionally, the research shows machine learning for mental health analysis to present a scalable method for wellbeing prediction. Integrating multiple data modalities, our model presents an integrative view of cognitive transformation that covers the gap between qualitative opinion and quantitative prediction. The contribution of this research is in presenting the viability of applying artificial intelligence (AI) models to facilitate enhanced mental health interventions through adaptive and personalized cognitive training programs. More generally, our results add to the emerging science of neuroplasticity-based cognitive training by delivering an evidence-based method for evaluating and predicting wellbeing improvement. The findings have implications that reach outside the research clinic, to clinical interventions, self-help programs, and mobile phone health applications, to offer a new mechanism for improving mental resilience and world life satisfaction through rigorous cognitive training.
- 14 October 2025
- Published by Multimedia University in Journal of Informatics and Web Engineering
- Vol. 4 (3) , 64-89
Open AccessTraining the Brain: A Machine Learning Approach to Predicting Wellbeing Through Intentional Thought Pattern Modification
This study provides a quantitative framework for wellbeing outcome prediction through intentional cognitive pattern alteration. We demonstrated 81.67% accurate prediction of wellbeing states, in a three-level classification (Low, Medium, High), using a Random Forest classifier with 16 features from psychological, physiological, and behavioural metrics. Our model singles out the gratitude cultivation (21.3%) and peace duration (23.7%) as the strongest predictors of positive well-being outcomes, which provides empirical support to traditional approaches of cognitive training with empirical evidence. Analysis of 1,000 synthetic cases shows that consistent practice of positive thought patterns over 3-6 months can strongly shift wellbeing states, with key behavioural markers showing progressive improvement which include increased joy moments, reduced anxiety episodes, and enhanced sleep quality. Our results establish that cognitive training outcomes can be quantitatively tracked and predicted with meaningful accuracy, hence providing a data-driven approach to mental health intervention design. Additionally, the research shows machine learning for mental health analysis to present a scalable method for wellbeing prediction. Integrating multiple data modalities, our model presents an integrative view of cognitive transformation that covers the gap between qualitative opinion and quantitative prediction. The contribution of this research is in presenting the viability of applying artificial intelligence (AI) models to facilitate enhanced mental health interventions through adaptive and personalized cognitive training programs. More generally, our results add to the emerging science of neuroplasticity-based cognitive training by delivering an evidence-based method for evaluating and predicting wellbeing improvement. The findings have implications that reach outside the research clinic, to clinical interventions, self-help programs, and mobile phone health applications, to offer a new mechanism for improving mental resilience and world life satisfaction through rigorous cognitive training.
- 1 October 2025
- Published by Global Academic Excellence (M) Sdn Bhd in International Journal of Modern Education
- Vol. 7 (27) , 692-705
AI-DRIVEN PERSONALIZED LEARNING PATHWAYS: TRANSFORMING EDUCATIONAL OUTCOMES THROUGH ADAPTIVE CONTENT DELIVERY SYSTEMS
This study presents a comprehensive AI-driven personalized learning system that transforms educational outcomes through adaptive content delivery mechanisms. We developed and evaluated an intelligent learning platform that leverages machine learning algorithms to create individualized learning pathways for diverse student populations. The platform utilizes student clustering, prediction algorithms of performance, and content recommendation systems to provide individualized learning experiences. Our empirical test with 500 students, 100 units of content, and 5,000 learning interactions shows significant educational outcome gains. The AI system attained 78.5% completion prediction accuracy and low prediction error of performance (MSE = 0.0112). Most importantly, students who utilized the personalized learning paths saw a mean 11.7% improvement in performance and a 6.3% increase in completion rates compared to the conventional means of learning. The platform successfully segmented four learner clusters, allowing for targeted intervention on underachieving students, average ones, high-achieving ones, and fast learners. Our findings provide compelling evidence that AI-driven personalization can forcefully counteract the inadequacies of one-size-fits-all education, with great potential for the advancement of educational equity and learning efficiency.
- 1 October 2025
- Published by Global Academic Excellence (M) Sdn Bhd in International Journal of Modern Education
- Vol. 7 (27) , 692-705
AI-DRIVEN PERSONALIZED LEARNING PATHWAYS: TRANSFORMING EDUCATIONAL OUTCOMES THROUGH ADAPTIVE CONTENT DELIVERY SYSTEMS
This study presents a comprehensive AI-driven personalized learning system that transforms educational outcomes through adaptive content delivery mechanisms. We developed and evaluated an intelligent learning platform that leverages machine learning algorithms to create individualized learning pathways for diverse student populations. The platform utilizes student clustering, prediction algorithms of performance, and content recommendation systems to provide individualized learning experiences. Our empirical test with 500 students, 100 units of content, and 5,000 learning interactions shows significant educational outcome gains. The AI system attained 78.5% completion prediction accuracy and low prediction error of performance (MSE = 0.0112). Most importantly, students who utilized the personalized learning paths saw a mean 11.7% improvement in performance and a 6.3% increase in completion rates compared to the conventional means of learning. The platform successfully segmented four learner clusters, allowing for targeted intervention on underachieving students, average ones, high-achieving ones, and fast learners. Our findings provide compelling evidence that AI-driven personalization can forcefully counteract the inadequacies of one-size-fits-all education, with great potential for the advancement of educational equity and learning efficiency.
- 26 August 2025
- Published by Akademia Baru Publishing in Journal of Advanced Research in Business and Management Studies
- Vol. 40 (1) , 72-82
Leveraging Artificial Intelligence to Enhance Strategic Information Systems: A Competitive Edge for the Logistics Industry
Artificial Intelligence (AI) is revolutionizing how strategic information systems (SIS) operate across industries by offering organizations new ways to gain a sustainable competitive advantage. This paper examines the integration of AI technologies such as machine learning, robotics, natural language processing (NLP), and predictive analytics into the logistics sector’s SIS. The study identifies specific applications of AI within primary logistics activities, including inbound and outbound logistics, warehousing, marketing and sales, and customer service by using Porter’s Value Chain framework. The paper adopts a thematic analysis approach based on a systematic literature review of 25 studies, revealing how AI enhances operational efficiency, cost reduction, and customer satisfaction. Findings indicate that AI enables superior delivery standards, optimized pricing, and smarter inventory and fleet management. The paper also highlights critical challenges such as ethical risks, data security, and system integration. Recommendations include implementing AI governance frameworks, ensuring GDPR compliance, and adopting phased integration strategies with expert support. This study offers practical and theoretical insights into the role of AI in strategic systems and emphasizes the importance of innovation balanced with risk management to gain a competitive edge in logistics.
- 26 August 2025
- Published by Akademia Baru Publishing in Journal of Advanced Research in Business and Management Studies
- Vol. 40 (1) , 72-82
Leveraging Artificial Intelligence to Enhance Strategic Information Systems: A Competitive Edge for the Logistics Industry
Artificial Intelligence (AI) is revolutionizing how strategic information systems (SIS) operate across industries by offering organizations new ways to gain a sustainable competitive advantage. This paper examines the integration of AI technologies such as machine learning, robotics, natural language processing (NLP), and predictive analytics into the logistics sector’s SIS. The study identifies specific applications of AI within primary logistics activities, including inbound and outbound logistics, warehousing, marketing and sales, and customer service by using Porter’s Value Chain framework. The paper adopts a thematic analysis approach based on a systematic literature review of 25 studies, revealing how AI enhances operational efficiency, cost reduction, and customer satisfaction. Findings indicate that AI enables superior delivery standards, optimized pricing, and smarter inventory and fleet management. The paper also highlights critical challenges such as ethical risks, data security, and system integration. Recommendations include implementing AI governance frameworks, ensuring GDPR compliance, and adopting phased integration strategies with expert support. This study offers practical and theoretical insights into the role of AI in strategic systems and emphasizes the importance of innovation balanced with risk management to gain a competitive edge in logistics.
- Open Access
- 4 April 2025
- Published by MDPI AG in International Journal of Environmental Research and Public Health
- Vol. 22 (4) , 562
Predictive Modeling for Pandemic Forecasting: A COVID-19 Study in New Zealand and Partner Countries
This study proposes a data-driven approach to leveraging large-scale COVID-19 datasets to enhance the predictive modeling of disease spread in the early stages. We systematically evaluate three machine learning models—ARIMA, Prophet, and LSTM—using a comprehensive framework that incorporates time-series analysis, multivariate data integration, and a Multi-Criteria Decision Making (MCDM) technique to assess model performance. The study focuses on key features such as daily confirmed cases, geographic variations, and temporal trends, while considering data constraints and adaptability across different scenarios. Our findings reveal that LSTM and ARIMA consistently outperform Prophet, with LSTM achieving the highest predictive accuracy in most cases, particularly when trained on 20-week datasets. ARIMA, however, demonstrates superior stability and reliability across varying time frames, making it a robust choice for short-term forecasting. A direct comparative analysis with existing approaches highlights the strengths and limitations of each model, emphasizing the importance of region-specific data characteristics and training periods. The proposed methodology not only identifies optimal predictive strategies but also establishes a foundation for automating predictive analysis, enabling timely and data-driven decision-making for disease control and prevention. This research is validated using data from New Zealand and its major trading partners—China, Australia, the United States, Japan, and Germany—demonstrating its applicability across diverse contexts. The results contribute to the development of adaptive forecasting frameworks that can empower public health authorities to respond proactively to emerging health threats.
- 4 April 2025
- Published by MDPI AG in International Journal of Environmental Research and Public Health
- Vol. 22 (4) , 562
Open AccessPredictive Modeling for Pandemic Forecasting: A COVID-19 Study in New Zealand and Partner Countries
This study proposes a data-driven approach to leveraging large-scale COVID-19 datasets to enhance the predictive modeling of disease spread in the early stages. We systematically evaluate three machine learning models—ARIMA, Prophet, and LSTM—using a comprehensive framework that incorporates time-series analysis, multivariate data integration, and a Multi-Criteria Decision Making (MCDM) technique to assess model performance. The study focuses on key features such as daily confirmed cases, geographic variations, and temporal trends, while considering data constraints and adaptability across different scenarios. Our findings reveal that LSTM and ARIMA consistently outperform Prophet, with LSTM achieving the highest predictive accuracy in most cases, particularly when trained on 20-week datasets. ARIMA, however, demonstrates superior stability and reliability across varying time frames, making it a robust choice for short-term forecasting. A direct comparative analysis with existing approaches highlights the strengths and limitations of each model, emphasizing the importance of region-specific data characteristics and training periods. The proposed methodology not only identifies optimal predictive strategies but also establishes a foundation for automating predictive analysis, enabling timely and data-driven decision-making for disease control and prevention. This research is validated using data from New Zealand and its major trading partners—China, Australia, the United States, Japan, and Germany—demonstrating its applicability across diverse contexts. The results contribute to the development of adaptive forecasting frameworks that can empower public health authorities to respond proactively to emerging health threats.
- 4 March 2025
- Published by Springer Nature
- p. 275-289
A Website Development of UCSI e-Market Hub: Transforming Unwanted Possessions
- 4 March 2025
- Published by Springer Nature
- p. 275-289
A Website Development of UCSI e-Market Hub: Transforming Unwanted Possessions