Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Anindya Nag has a total of 165 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Exploring the Power of Machine Learning in Analysing Protein–Protein Sequences

    by Anindya Nag, Riya Sil, Mehedi Hassan, Tamanna Haque Ritu, Biva Das, Nilanjana Roy, Anurag Sinha, Dron Guin, Nishchal Maurya

    Proteins are fundamental biological macromolecules responsible for regulating nearly all cellular processes, and their functions are largely determined by the underlying amino acid sequences. Understanding the relationship between protein sequences and their structural and functional properties remains a major challenge in molecular biology. Although experimental techniques provide accurate insights, they are often time‐consuming and expensive, which has led to increasing reliance on computational approaches for protein sequence analysis and structure prediction. Numerous techniques—including homology modelling, machine learning, deep learning, natural language processing and other artificial intelligence methods—have been developed for analysing protein sequences and predicting protein–protein interactions. However, the rapidly growing number of computational methods makes it difficult for researchers to systematically evaluate and select suitable approaches for specific biological applications. This study presents a comprehensive review of recent computational techniques for protein sequence analysis, focusing on machine learning‐ and deep learning‐based frameworks used for predicting protein structure and interactions. This review systematically categorises existing approaches based on their methodological foundations, datasets and performance characteristics and provides a comparative discussion of their advantages and limitations. Furthermore, this study highlights current research gaps, challenges in large‐scale protein analysis and emerging trends in AI‐driven protein modelling. The findings of this review provide a structured reference for researchers and practitioners, facilitating the selection of appropriate computational techniques and supporting future advancements in protein sequence analysis, disease research and drug discovery.


  • Transforming Global Healthcare

    by Anindya Nag, Mehedi Hassan, Riya Sil, Asif Karim

    Transforming Global Healthcare: The Role of AI and IoT in Modern Medical Systems explores the transformative potential of the integration of technologies such as AI, Big Data, IoT, and Blockchain in the healthcare sector. This comprehensive volume aims to illustrate how these advanced technologies can address sustainability challenges, improve decision-making processes, and enhance overall healthcare delivery. By mapping existing knowledge and forecasting future trends, the book provides a detailed roadmap for sustainable healthcare development. Key features include: – Explores AI, big data, IoT, and blockchain in healthcare in-depth. – Provides actionable strategies for energy management, waste disposal, and green supply chains – Provides in-depth coverage of real-time health monitoring, predictive analytics, robotic surgery, and AI medical imaging. – Provides insights into preventing drug counterfeiting, managing pharmaceutical cold chains, and optimizing supply chains. – Comprises a comparative analysis of global healthcare systems, telemedicine evolution, and future digital health trends.


  • The Generative AI Revolution in Neurodegenerative Disease Research

    by Ayontika Das, Hirak Mondal, Teresa Jency Bala, Akash Adhikary, Anindya Nag

    NDs, including Alzheimer’s disease and Parkinson’s disease, are chronic conditions marked by progressive neuronal dysfunction and cell death. Diagnosing NDs remains a complex challenge, as it necessitates the integration of diverse biomarker data, such as genetic, proteomic, and neuroimaging information, which exhibit multidimensional, non-linear relationships. The emergence of AI technologies, particularly ML and DL algorithms, has introduced innovative methodologies to address these diagnostic complexities. By leveraging AI-driven tools, researchers and clinicians can extract critical patterns and insights from large-scale, heterogeneous datasets, enabling more accurate, timely, and objective clinical evaluations. This review examines the current applications of AI in ND diagnosis, evaluates its contributions to enhancing diagnostic precision and efficiency, and discusses persistent limitations and challenges. Furthermore, the authors propose potential directions for future research to optimize AI’s role in advancing ND diagnostics.

  • Unlocking Neural Pathways

    by Teresa Jency Bala, Hirak Mondal, Akash Adhikary, Anindya Nag

    Detection and treatment of brain disorders such as Alzheimer’s disease, Parkinson’s disease, and other forms of dementia can be challenging. This is due to the fact that they affect people differently, progress gradually, and begin with subtle symptoms. Recent advances in large language models (LLMs) and their architectures exemplified by GPT and BERT offer capabilities for solving these problems across all domains of neurological care. This chapter discusses how LLMs introduce new paradigms to the treatment of neurodegenerative diseases through the myriad applications of LLMs to early detection, modeling disease progression, cognitive assessment, rehabilitation, and caregiver support. LLMs extract latent cognitive attributes from spontaneous speech, predict the onset of diseases using linguistic biomarkers, and implement high-accuracy interpretation of neuropsychological tests. This understanding allows the development of comprehensive digital biomarkers infused with neuroimaging data, wearable sensors, and electronic health records for tracking the trajectories of diseases and treatment responses. We delve into LLM considerations for personalized therapy, from dynamic cognitive stimulation tasks to adaptable pace analysis, which provides real-time feedback for individualized care strategies. Ethical discussions notably include data privacy, model bias, and transparency, which leads to outlining clinical validation pathways for real-world paperwork. Through bridging deep learning, natural language processing, and medical knowledge, LLMs provide a roadmap to a future in which AI-augmented neurology enables early interventions, accurate diagnoses, and improved quality of life for patients and their caregivers. This chapter is both a technical guide and a visionary blueprint that will guide the integration of LLMs into the neurological healthcare ecosystem.


  • Open Access

    Advancing Federated Learning: A Systematic Literature Review of Methods, Challenges, and Applications

    by Tamanna Zubairi Sana, Shahab Abdulla, Anindya Nag, Ayontika Das, Mehedi Hassan, Zoya Zubairi Fiza, Asif Karim, Sheikh Ridwan Raihan Kabir

    Federated Learning (FL) has emerged as a cutting-edge paradigm in machine learning, showcasing remarkable advancements in recent years. This research paper delves into the dynamic landscape of FL by addressing four pivotal research questions. The study investigates the most recent advancements in implementing FL and explores additional applications that could benefit from this decentralized learning paradigm. This inquiry aims to provide an up-to-date overview of the evolving FL field and its potential cross-industry impact. The paper explores the integration of FL with various machine learning approaches to ensure optimal performance, privacy preservation, and scalability. By unraveling the collaborative aspects of FL with other machine learning paradigms, the research seeks to unveil novel strategies for enhancing efficiency in FL scenarios. The third research question focuses on the repercussions of scalability challenges and resource constraints in federated learning. This investigation aims to uncover the practical difficulties of implementing FL across diverse sectors, shedding light on potential barriers to its widespread adoption. The research probes into the future of federated learning by examining how it will be utilized in upcoming technological advancements and industries. This exploration aims to provide insights into the long-term viability and applicability of FL, anticipating its role in shaping the technological landscape across various sectors. Through a comprehensive analysis of these research questions, this paper contributes to the understanding of FL, providing valuable insights for researchers, practitioners, and decision-makers navigating the intricate intersection of FL, machine learning, and emerging technologies. This research paper aspires to provide a holistic overview of the advancements, integration possibilities, challenges, and prospects associated with federated learning, contributing to the ongoing discourse on the intersection of FL and machine learning in contemporary technological landscapes.


  • Enhancing IoT security

    by Irin Sultana, Hirak Mondal, Azizul Islam, Swan Mollick, Teresa Jency Bala, Anindya Nag

    While the Internet of Things (IoT) enables the interconnection and sharing of data across many devices, it also exposes them to potential security vulnerabilities. This chapter addresses the previously described security concerns by utilizing machine learning (ML) techniques to recognize and identify potential risks and anomalies in IoT devices. We utilize the ML-Edge IIoT dataset, which merges data from diverse IoT devices and covers 14 distinct threat categories along with benign network traffic. We employ ML techniques, including Random Forest (RF), Extreme Gradient Boosting (XGBoost), K-Nearest Neighbours (KNN), Categorical Boosting (CatBoost), Decision Trees (DTs), and Naive Bayes (NB). CatBoost and XGBoost obtained an astounding accuracy rate of up to 95%. The RF, DT, Logistic Regression, KNN, and NB models exhibit accuracy rates of 94%, 93%, 89%, 87%, and 85%, respectively. The primary aim of this chapter is to enhance the security of the IoT by identifying various types of assaults through the assessment of model performance. Our research findings demonstrate that ML possesses a high level of proficiency in accurately identifying and mitigating security vulnerabilities in IoT networks. The chapter offers a robust and dependable solution to the increasing prevalence of cyberattacks that specifically target IoT devices.

  • Navigating Autism Spectrum Disorder: A Fusion of Deep Learning and Explainable AI for Enhanced Detection and Classification

    by K. M. Danial Sadad, Anindya Nag, Nitesh Upadhyaya, Salman Islam, Kaysun Molla, Dip Ranjan Dutta

    Autism spectrum disorder (ASD) is a neurodevelopmental disease that impairs communication, conduct, and social interactions. Early discovery is crucial for effective management. This study investigates the application of deep learning algorithms to predict ASD in children using a freshly constructed dataset. Several models were tested, including feed forward neural networks (FFNN), deep neural networks (DNN), convolutional neural networks (CNN), long short-term memory (LSTM), and gated recurrent units (GRU), with the DNN model outperforming the others with a 99.22% accuracy. Furthermore, this study employs explainable AI, namely local interpretable model-agnostic explanations (LIME), to give insights into model predictions. Explain ability integration guarantees that forecasts are interpretable, which improves model confidence and transparency. This study 166contributes a valuable dataset and demonstrates the efficacy of deep learning approaches and explainable AI in early ASD detection.


  • Advanced Fingerprint Authentication System and Neurodegenerative Disorder Multi-modal Pattern Recognition Techniques Using Deep Learning

    by Sajol Debnath, K. M. Danial Sadad, Shamim Parvej, Shaik Abdul Kareem, Kaysun Molla, Anindya Nag

    Biometric fingerprint identification systems are critical for improving security across a variety of businesses; nevertheless, single biometric approaches frequently have limits due to non-universality and noise, resulting in lower genuine acceptance rates (GAR) and greater false acceptance rates (FAR). To overcome these issues, this work investigates a multimodal biometric system that integrates fingerprint data with other biometric traits to improve accuracy and resistance to spoofing. We create a multimodal pattern recognition framework with deep learning models like EfficientNet (89%), WideResNet (94%), Xception (83%), DenseNet (91%), and CNN (88%), with WideResNet attaining the highest classification accuracy. Beyond security, generative AI applications within this framework have the potential to advance neurodegenerative illness research by allowing for the analysis of biometric data to diagnose and monitor cognitive decline early on. This method dramatically improves biometric systems’ resilience, security, and dependability, making it appropriate for both high-security situations and noninvasive neuropsychological examinations.


  • Psychological Stress Prediction of Human Using Machine Learning Algorithms

    by M. Raihan, Anindya Nag, Mehedi Hassan, Anupam Kumar Bairagi

    Psychological Stress is the most common health issue nowadays. Stress in the workplace poses significant risks to workers' health and hinders an organization's ability to function efficiently. Depression is a widespread disorder of the brain and nervous system. This study aims to forecast psychological stress levels utilizing the Burns Depression Checklist (BDC). In this study, four distinct machine learning methodologies has employed, specifically AdaBoost, Bagging, K-Nearest Neighbors (KNN), and Stochastic Gradient Descent (SGD) for predicting psychological stress in this study. According to the analysis, the KNN classifier with Genetic Search and Ranker feature selection techniques demonstrates a significantly higher level of efficacy compared to other approaches. KNN classifier provides satisfactory outcomes with both 13 and 15 selected features. The Shapley Additive Explanations (SHAP) analysis justifies the outcome.

  • Explainable artificial intelligence for natural language processing: A survey

    by Mehedi Hassan, Anindya Nag, Riya Biswas, Shahin Ali, Sadika Zaman, Anupam Kumar Bairagi, Chetna Kaushal