Shefayatuj Johara Chowdhury
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Shefayatuj Johara Chowdhury has a total of 39 co-authors. Here are the top co-authors by number of shared publications:
- Farzana Tasnim
International Islamic University Chittagong
4 shared publications - Mohammad Shahadat Hossain
University of Chittagong
4 shared publications - Tanjim Mahmud
Department of Computer Science and Engineering, Rangamati Science and Technology University, Rangamati 4500, Bangladesh
4 shared publications - Abubokor Hanip
Washington University of Science and Technology
2 shared publications - Faisal Bin Al Abid
Universiti Teknologi Malaysia
2 shared publications - Jia Uddin
Endicott College, Woosong University
2 shared publications
Latest Publications
- Open Access
- 1 June 2026
- Published by Institute of Advanced Engineering and Science in International Journal of Advances in Applied Sciences
- Vol. 15 (2) , 479-489
Predicting Indonesian academician turnover intention: validity and reliability analysis
This study evaluates Indonesian academic turnover intention (TOI) by analyzing demographic and work-related factors through feature selection methods and utilizes random forest (RF) as a baseline classifier for TOI prediction, while applying statistical methods to ensure the reliability of the collected primary dataset. The main advantage of this approach is to find out the importance of these factors with statistical validation to reliably investigate Indonesian academicians’ TOI. Feature selection methods such as information gain (IG) and SelectKBest were used to find out feature importance, while the reliability of the dataset was assessed through statistical approaches such as Cronbach alpha, confirmatory factor analysis (CFA), average variance extracted (AVE), and consistency ratio (CR). To test the importance of demographic and work-related factors, Python was used as an implementation tool for the Indonesian academic TOI dataset (IRB reference: 19.12.4/UN32.14/PB/2024), comprising 527 samples. The superiority of the importance of work-related factors in contrast to demographic factors was consistently demonstrated by feature selection methods, and a statistical approach confirmed the reliability of the collected primary dataset, consequently ensuring the robustness of the findings. It is envisaged that this approach can be very useful for human resource (HR) departments to pay more attention to the important demographic factors for reducing Indonesian academic TOI.
- 1 June 2026
- Published by Institute of Advanced Engineering and Science in International Journal of Advances in Applied Sciences
- Vol. 15 (2) , 479-489
Open AccessPredicting Indonesian academician turnover intention: validity and reliability analysis
This study evaluates Indonesian academic turnover intention (TOI) by analyzing demographic and work-related factors through feature selection methods and utilizes random forest (RF) as a baseline classifier for TOI prediction, while applying statistical methods to ensure the reliability of the collected primary dataset. The main advantage of this approach is to find out the importance of these factors with statistical validation to reliably investigate Indonesian academicians’ TOI. Feature selection methods such as information gain (IG) and SelectKBest were used to find out feature importance, while the reliability of the dataset was assessed through statistical approaches such as Cronbach alpha, confirmatory factor analysis (CFA), average variance extracted (AVE), and consistency ratio (CR). To test the importance of demographic and work-related factors, Python was used as an implementation tool for the Indonesian academic TOI dataset (IRB reference: 19.12.4/UN32.14/PB/2024), comprising 527 samples. The superiority of the importance of work-related factors in contrast to demographic factors was consistently demonstrated by feature selection methods, and a statistical approach confirmed the reliability of the collected primary dataset, consequently ensuring the robustness of the findings. It is envisaged that this approach can be very useful for human resource (HR) departments to pay more attention to the important demographic factors for reducing Indonesian academic TOI.
- 1 March 2026
- Published by IOP Publishing in Engineering Research Express
- Vol. 8 (5) , 055214
A hybrid deep learning framework for identifying Aedes aegypti mosquitoes using CNN, ViT, and GNN architectures
Dengue has remained a major global health concern, increasing the need for accurate and scalable mosquito surveillance systems, particularly for identifying Aedes aegypti, the primary vector of dengue transmission. To address this challenge, a hybrid deep learning framework, MC-TransNet, has been developed to enhance mosquito species classification. The framework has integrated four complementary architectures—Big Transfer ResNet50x1 (BiT), DenseNet121, Vision Transformer (ViT), and Graph Neural Networks (GNN)—allowing both fine-grained morphological details and long-range structural information to be captured effectively.The model has been trained using a dataset of 2,640 images representing three medically significant mosquito species. MC-TransNet has achieved 98.55% accuracy, 93.10% precision, 100% recall, 96.42% F1-score, and a Matthews Correlation Coefficient (MCC) of 96.96%, demonstrating strong and balanced predictive performance across all classes. These results have surpassed the performance reported in earlier CNN-based and transformer-based studies.To support transparency and field applicability, explainable AI techniques—Grad-CAM, Grad-CAM++, and LIME—have been incorporated to highlight key anatomical regions influencing classification decisions. Overall, the findings have indicated that MC-TransNet offers a robust and interpretable solution for automated mosquito identification, with significant potential for strengthening real-time vector surveillance and dengue prevention efforts.
- 1 March 2026
- Published by IOP Publishing in Engineering Research Express
- Vol. 8 (5) , 055214
A hybrid deep learning framework for identifying Aedes aegypti mosquitoes using CNN, ViT, and GNN architectures
Dengue has remained a major global health concern, increasing the need for accurate and scalable mosquito surveillance systems, particularly for identifying Aedes aegypti, the primary vector of dengue transmission. To address this challenge, a hybrid deep learning framework, MC-TransNet, has been developed to enhance mosquito species classification. The framework has integrated four complementary architectures—Big Transfer ResNet50x1 (BiT), DenseNet121, Vision Transformer (ViT), and Graph Neural Networks (GNN)—allowing both fine-grained morphological details and long-range structural information to be captured effectively.The model has been trained using a dataset of 2,640 images representing three medically significant mosquito species. MC-TransNet has achieved 98.55% accuracy, 93.10% precision, 100% recall, 96.42% F1-score, and a Matthews Correlation Coefficient (MCC) of 96.96%, demonstrating strong and balanced predictive performance across all classes. These results have surpassed the performance reported in earlier CNN-based and transformer-based studies.To support transparency and field applicability, explainable AI techniques—Grad-CAM, Grad-CAM++, and LIME—have been incorporated to highlight key anatomical regions influencing classification decisions. Overall, the findings have indicated that MC-TransNet offers a robust and interpretable solution for automated mosquito identification, with significant potential for strengthening real-time vector surveillance and dengue prevention efforts.
- 10 December 2025
- Published by IOP Publishing in Engineering Research Express
- Vol. 7 (4) , 0452c9
Analysis of Alzheimer’s disease progression stages using machine learning
The early detection and classification of Alzheimer’s Disease (AD) present substantial challenges due to the complex and subtle nature of its progression. Traditional diagnostic methods are invasive, costly, and time-consuming, underscoring the need for efficient alternatives. To address the challenges, this research proposes an approach the ADNI TADPOLE benchmark dataset and evaluate multiple machine learning algorithms—K-Nearest Neighbors, Support Vector Machine, Naïve Bayes, CatBoost, and Gradient Boosting—trained on an integrated feature set combining imaging-derived volumetric measures with cognitive and clinical markers. Gradient Boosting achieved the highest performance, with 98.98% accuracy and an MCC of 0.97, consistently outperforming other methods. Unlike prior studies emphasizing imaging alone, this work highlights the predictive value of multimodal data integration for AD stage classification. These findings demonstrate that Gradient Boosting provides a robust and clinically translatable framework for timely AD detection and personalized treatment planning.
- 10 December 2025
- Published by IOP Publishing in Engineering Research Express
- Vol. 7 (4) , 0452c9
Analysis of Alzheimer’s disease progression stages using machine learning
The early detection and classification of Alzheimer’s Disease (AD) present substantial challenges due to the complex and subtle nature of its progression. Traditional diagnostic methods are invasive, costly, and time-consuming, underscoring the need for efficient alternatives. To address the challenges, this research proposes an approach the ADNI TADPOLE benchmark dataset and evaluate multiple machine learning algorithms—K-Nearest Neighbors, Support Vector Machine, Naïve Bayes, CatBoost, and Gradient Boosting—trained on an integrated feature set combining imaging-derived volumetric measures with cognitive and clinical markers. Gradient Boosting achieved the highest performance, with 98.98% accuracy and an MCC of 0.97, consistently outperforming other methods. Unlike prior studies emphasizing imaging alone, this work highlights the predictive value of multimodal data integration for AD stage classification. These findings demonstrate that Gradient Boosting provides a robust and clinically translatable framework for timely AD detection and personalized treatment planning.
- Open Access
- 8 November 2025
- Published by MDPI AG in Diagnostics
- Vol. 15 (22) , 2833
A Hybrid MCDM and Machine Learning Framework for Thalassemia Risk Assessment in Pregnant Women
Background: Thalassemia has been recognized as a critical public health issue in Bangladesh, especially among pregnant women, due to its hereditary nature and the lack of early screening infrastructure. Early identification of at-risk individuals is essential to prevent the transmission of this genetic disorder to future generations and to reduce the burden on an already strained healthcare system. Methods: In this study, an innovative framework for thalassemia risk assessment has been developed by integrating Multi-Criteria Decision-Making (MCDM) methods—specifically AHP-TOPSIS—with machine learning algorithms including Random Forest, XGBoost, and CatBoost. Explainable Artificial Intelligence (XAI) techniques such as SHAP and LIME have also been incorporated to improve model transparency and trustworthiness. Real-world clinical and demographic data, consisting of 16 features and 1200 samples, have been collected through a structured survey and processed using rigorous feature selection and ranking methods. Risk stratification has been performed to classify patients into high, medium, and low categories, enabling targeted intervention. Results: Among all models, the XGBoost classifier trained on AHP–TOPSIS–prioritized features achieved a consistent accuracy of 99.28% under stratified 20-fold cross-validation, demonstrating robust diagnostic classification performance. The model predominantly captures hematologic patterns characteristic of thalassemia manifestations, functioning as an assistive diagnostic framework rather than a causal risk predictor. The explainability of predictions, ensured through comprehensive visual and statistical analyses, further enhances the model’s clinical transparency and reliability. Conclusions: The proposed MCDM–machine learning framework demonstrates strong potential for improving thalassemia risk assessment, enabling early detection and informed decision-making in maternal healthcare. The proposed framework should be regarded as a preliminary proof-of-concept system that demonstrates the feasibility of integrating Multi-Criteria Decision-Making (AHP–TOPSIS) with advanced machine learning and explainable-AI techniques for thalassemia assessment. Although the model achieved strong diagnostic performance under nested cross-validation, additional external validation and inclusion of causal predictors are required before clinical deployment.
- 8 November 2025
- Published by MDPI AG in Diagnostics
- Vol. 15 (22) , 2833
Open AccessA Hybrid MCDM and Machine Learning Framework for Thalassemia Risk Assessment in Pregnant Women
Background: Thalassemia has been recognized as a critical public health issue in Bangladesh, especially among pregnant women, due to its hereditary nature and the lack of early screening infrastructure. Early identification of at-risk individuals is essential to prevent the transmission of this genetic disorder to future generations and to reduce the burden on an already strained healthcare system. Methods: In this study, an innovative framework for thalassemia risk assessment has been developed by integrating Multi-Criteria Decision-Making (MCDM) methods—specifically AHP-TOPSIS—with machine learning algorithms including Random Forest, XGBoost, and CatBoost. Explainable Artificial Intelligence (XAI) techniques such as SHAP and LIME have also been incorporated to improve model transparency and trustworthiness. Real-world clinical and demographic data, consisting of 16 features and 1200 samples, have been collected through a structured survey and processed using rigorous feature selection and ranking methods. Risk stratification has been performed to classify patients into high, medium, and low categories, enabling targeted intervention. Results: Among all models, the XGBoost classifier trained on AHP–TOPSIS–prioritized features achieved a consistent accuracy of 99.28% under stratified 20-fold cross-validation, demonstrating robust diagnostic classification performance. The model predominantly captures hematologic patterns characteristic of thalassemia manifestations, functioning as an assistive diagnostic framework rather than a causal risk predictor. The explainability of predictions, ensured through comprehensive visual and statistical analyses, further enhances the model’s clinical transparency and reliability. Conclusions: The proposed MCDM–machine learning framework demonstrates strong potential for improving thalassemia risk assessment, enabling early detection and informed decision-making in maternal healthcare. The proposed framework should be regarded as a preliminary proof-of-concept system that demonstrates the feasibility of integrating Multi-Criteria Decision-Making (AHP–TOPSIS) with advanced machine learning and explainable-AI techniques for thalassemia assessment. Although the model achieved strong diagnostic performance under nested cross-validation, additional external validation and inclusion of causal predictors are required before clinical deployment.
- Open Access
- 1 September 2025
- Published by MDPI AG in Technologies
- Vol. 13 (9) , 385
Predicting Graduate Employability Using Hybrid AHP-TOPSIS and Machine Learning: A Moroccan Case Study
The persistent issue of unemployment and the mismatch between graduate skills and labor market demands has drawn increasing attention from academics and educational institutions, especially in light of rapid advancements in technology. Emerging technologies such as Artificial Intelligence (AI) and Machine Learning (ML) offer valuable opportunities to analyze job market dynamics. In this work, we present a novel framework aimed at predicting graduate employability using current labor market data from Morocco. Our approach combines Multi-Criteria Decision-Making (MCDM) techniques with ML-based predictive models. AHP prioritizes employability factors and TOPSIS ranks skill demands—together forming input features for machine learning models. 2100 job listings obtained through web scraping, we trained and evaluated several ML models. Among them, the K-Nearest Neighbors (KNN) classifier demonstrated the highest accuracy, achieving 99.71% accuracy through 5-fold cross-validation. While the study is based on a limited dataset, it highlights the practical relevance of combining MCDM methods with ML for employability prediction. This study is among the first to integrate AHP–TOPSIS with KNN for employability prediction using real-time Moroccan labor market data.
- 1 September 2025
- Published by MDPI AG in Technologies
- Vol. 13 (9) , 385
Open AccessPredicting Graduate Employability Using Hybrid AHP-TOPSIS and Machine Learning: A Moroccan Case Study
The persistent issue of unemployment and the mismatch between graduate skills and labor market demands has drawn increasing attention from academics and educational institutions, especially in light of rapid advancements in technology. Emerging technologies such as Artificial Intelligence (AI) and Machine Learning (ML) offer valuable opportunities to analyze job market dynamics. In this work, we present a novel framework aimed at predicting graduate employability using current labor market data from Morocco. Our approach combines Multi-Criteria Decision-Making (MCDM) techniques with ML-based predictive models. AHP prioritizes employability factors and TOPSIS ranks skill demands—together forming input features for machine learning models. 2100 job listings obtained through web scraping, we trained and evaluated several ML models. Among them, the K-Nearest Neighbors (KNN) classifier demonstrated the highest accuracy, achieving 99.71% accuracy through 5-fold cross-validation. While the study is based on a limited dataset, it highlights the practical relevance of combining MCDM methods with ML for employability prediction. This study is among the first to integrate AHP–TOPSIS with KNN for employability prediction using real-time Moroccan labor market data.
- Open Access
- 5 May 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 13, 83765-83780
Elevating E-Service Standards in Bangladesh: Merging FAHP and TOPSIS for Strategic Online Business Solutions
The rapid growth of Bangladesh in the digital market has made it more difficult to optimize the quality of e-services. The quick growth of online businesses makes it necessary to identify the critical elements that affect their performance. In contrast to other research, this study takes a novel approach by examining these issues from the viewpoint of business owners rather than consumers.According to a survey of the literature, the majority of studies have concentrated on the viewpoints of customers, with little attention paid to the difficulties faced by business owners. In order to close this gap, 57 online business owners participated in a survey that gathered information on ten important aspects, such as website architecture, information sharing, delivery methods, responsiveness, customization, customer service, dependability, payment security, return policies, and issues with service quality.The Random Forest Classifier was utilized to extract features from the data, and fuzzy triangular matrices were employed for pairwise comparisons. the Fuzzy Analytical Hierarchy Process (FAHP) and TOPSIS were used for decisionmaking and ranking. The results show that issues with service quality have the biggest influence on internet firms’ success. The model’s accuracy was also assessed using the XGBoost approach, which produced an astounding 93% accuracy and 93% F1-score, which is a notable improvement over previous studies.This study offers a methodical and thorough approach to evaluating the crucial success elements for internet businesses, highlighting the significance of payment security, service quality concerns, and other significant obstacles.
- 5 May 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 13, 83765-83780
Open AccessElevating E-Service Standards in Bangladesh: Merging FAHP and TOPSIS for Strategic Online Business Solutions
The rapid growth of Bangladesh in the digital market has made it more difficult to optimize the quality of e-services. The quick growth of online businesses makes it necessary to identify the critical elements that affect their performance. In contrast to other research, this study takes a novel approach by examining these issues from the viewpoint of business owners rather than consumers.According to a survey of the literature, the majority of studies have concentrated on the viewpoints of customers, with little attention paid to the difficulties faced by business owners. In order to close this gap, 57 online business owners participated in a survey that gathered information on ten important aspects, such as website architecture, information sharing, delivery methods, responsiveness, customization, customer service, dependability, payment security, return policies, and issues with service quality.The Random Forest Classifier was utilized to extract features from the data, and fuzzy triangular matrices were employed for pairwise comparisons. the Fuzzy Analytical Hierarchy Process (FAHP) and TOPSIS were used for decisionmaking and ranking. The results show that issues with service quality have the biggest influence on internet firms’ success. The model’s accuracy was also assessed using the XGBoost approach, which produced an astounding 93% accuracy and 93% F1-score, which is a notable improvement over previous studies.This study offers a methodical and thorough approach to evaluating the crucial success elements for internet businesses, highlighting the significance of payment security, service quality concerns, and other significant obstacles.
- 1 January 2025
- Published by Springer Nature
- p. 190-204
Improving Missing Data Imputation with GF-WAI: An Explainable AI-Based Ensemble Method for Dengue Disease Prediction
- 1 January 2025
- Published by Springer Nature
- p. 190-204
Improving Missing Data Imputation with GF-WAI: An Explainable AI-Based Ensemble Method for Dengue Disease Prediction
- 1 January 2025
- Published by Springer Nature
- p. 101-114
Detecting AI-Generated Text Using Machine Learning: A Comparative Study of MLP and RoBERTa-Based Models
- 1 January 2025
- Published by Springer Nature
- p. 101-114
Detecting AI-Generated Text Using Machine Learning: A Comparative Study of MLP and RoBERTa-Based Models
- 30 March 2024
- Published by Springer Nature
- p. 335-347
An Expert System to Monitor and Risk Assessment of Chronic Disease Patients Using FTOPSIS
- 30 March 2024
- Published by Springer Nature
- p. 335-347
An Expert System to Monitor and Risk Assessment of Chronic Disease Patients Using FTOPSIS
- Open Access
- 25 February 2024
- Published by College of Science for Women, University of Baghdad in Baghdad Science Journal
- Vol. 21 (2(SI)) , 0706
Simplified Novel Approach for Accurate Employee Churn Categorization using MCDM, De-Pareto Principle Approach, and Machine Learning
يعد طرد الموظفين من المنظمات مشكلة خطيرة. يجب حل مشكلة تدوير الموظفين أو اطردهم من داخل المنظمة نظرًا لأن لها تأثيرًا سلبيًا على المنظمة. يعد الاكتشاف اليدوي لتسرب الموظفين أمرًا صعبًا للغاية، لذلك تم استخدام خوارزميات التعلم الآلي ML بشكل متكرر لاكتشاف تسرب الموظفين بالإضافة إلى تصنيف الموظفين وفقًا للاستبدالهم. باستخدام التعلم الآلي، بحثت دراسة واحدة فقط في تصنيف الموظفين حتى الآن. تم اقتراح نهج جديد لاتخاذ القرار متعدد المعايير MCDM إلى جانب مبدأ DE-PARETO لتصنيف الموظفين. يشار إلى هذا باسم مخطط SNEC . تم تصميم نموذج AHP-TOPSIS DE-PARETO PRINCIPLE (AHPTOPDE) الذي يستخدم نظام MCDM على مرحلتين لتصنيف الموظفين. في المرحلة الأولى، تم استخدام عملية التسلسل الهرمي التحليلي AHP لتعيين الأوزان النسبية لعوامل إنجاز الموظف. في المرحلة الثانية، تم استخدام TOPSIS للتعبير عن أهمية الموظفين في إجراء تصنيف الموظفين. تم تطبيق قاعدة 20-30-50 البسيطة في مبدأ DE PARETO لتصنيف الموظفين إلى ثلاث مجموعات رئيسية وهي الموظفون المتحمسون والسلوكيون والمضطربون. يتم بعد ذلك تطبيق خوارزمية الغابة العشوائية كخوارزمية أساسية لإطار عمل الموظفين المقترح للتنبؤ بخسارة الموظفين على أساس الفصل...
- 25 February 2024
- Published by College of Science for Women, University of Baghdad in Baghdad Science Journal
- Vol. 21 (2(SI)) , 0706
Open AccessSimplified Novel Approach for Accurate Employee Churn Categorization using MCDM, De-Pareto Principle Approach, and Machine Learning
يعد طرد الموظفين من المنظمات مشكلة خطيرة. يجب حل مشكلة تدوير الموظفين أو اطردهم من داخل المنظمة نظرًا لأن لها تأثيرًا سلبيًا على المنظمة. يعد الاكتشاف اليدوي لتسرب الموظفين أمرًا صعبًا للغاية، لذلك تم استخدام خوارزميات التعلم الآلي ML بشكل متكرر لاكتشاف تسرب الموظفين بالإضافة إلى تصنيف الموظفين وفقًا للاستبدالهم. باستخدام التعلم الآلي، بحثت دراسة واحدة فقط في تصنيف الموظفين حتى الآن. تم اقتراح نهج جديد لاتخاذ القرار متعدد المعايير MCDM إلى جانب مبدأ DE-PARETO لتصنيف الموظفين. يشار إلى هذا باسم مخطط SNEC . تم تصميم نموذج AHP-TOPSIS DE-PARETO PRINCIPLE (AHPTOPDE) الذي يستخدم نظام MCDM على مرحلتين لتصنيف الموظفين. في المرحلة الأولى، تم استخدام عملية التسلسل الهرمي التحليلي AHP لتعيين الأوزان النسبية لعوامل إنجاز الموظف. في المرحلة الثانية، تم استخدام TOPSIS للتعبير عن أهمية الموظفين في إجراء تصنيف الموظفين. تم تطبيق قاعدة 20-30-50 البسيطة في مبدأ DE PARETO لتصنيف الموظفين إلى ثلاث مجموعات رئيسية وهي الموظفون المتحمسون والسلوكيون والمضطربون. يتم بعد ذلك تطبيق خوارزمية الغابة العشوائية كخوارزمية أساسية لإطار عمل الموظفين المقترح للتنبؤ بخسارة الموظفين على أساس الفصل...