Syed Ziaur Rahman
Research Subjects
Syed Ziaur Rahman has a total of 36 co-authors. Here are the top co-authors by number of shared publications:
- Akram Pasha
Fujairah University
5 shared publications - D. N. Puneeth Kumar
Atria Institute of Technology
3 shared publications - Shaik Sayeed Ahamed
Atria Institute of Technology
3 shared publications - Khaja Mannanuddin
Department of Computer Science and Engineering, School of CS&AI, SR University, Hasanparthy, Warangal, Telangana, India
2 shared publications - Afifa Salsabil Fathima
School of Computer Science and Engineering, REVA University, Bengaluru, Karnataka, India
1 shared publication - Amarendra Kothalanka
Department of Computer Science and Engineering, Koneru Lakshmaiah Education Foundation, Vaddeswaram 522502, Andhra Pradesh, India
1 shared publication
Latest Publications
- 12 November 2026
- Published by World Scientific Pub Co Pte Ltd in Journal of Circuits, Systems and Computers
Automated Detection of Alzheimer’s Disease from EEG Signals Using VMD-Driven Multi-Domain Features and Rich and Poor Optimization-Enhanced Classification
Alzheimer’s disease is a progressive neurodegenerative disorder that affects large populations worldwide and produces measurable alterations in brain electrical activity. In this study, we propose an automated EEG-based methodology for distinguishing Alzheimer’s disease, mild cognitive impairment (MCI) and healthy control subjects. The EEG signals are first decomposed using adaptive variational mode decomposition (VMD) to obtain narrow-band intrinsic modes that capture key nonlinear and nonstationary properties of the recordings. A comprehensive set of time-domain, spectral, Hilbert envelope, and permutation entropy features is then extracted from each mode, resulting in an initial feature pool of 56 descriptors. These features are reduced using the rich and poor optimization (RPO) algorithm, and their discriminative ability is assessed using the Kruskal-Wallis statistical test with Bonferroni correction applied to control the family-wise error rate across all 56 features (corrected threshold [Formula: see text]). Forty-five statistically significant features are retained after Bonferroni correction and used as the final input feature set for classification. Three classical classifiers, namely random forest (RF), K-nearest neighbor (KNN) and support vector machine (SVM), are trained using 10-fold cross-validation with the synthetic minority oversampling technique (SMOTE) applied strictly within each training fold to prevent data leakage in order to evaluate the proposed framework. The RF model achieves the highest performance with a classification accuracy of 82.10%, a balanced accuracy of 71.90%, a macro-F1-score of 0.6534 and a macro area under the ROC curve (AUC) of 0.8933, supported by high sensitivity, specificity and predictive values. Cross-dataset validation on an independent publicly available EEG dataset (OpenNeuro ds004504, 36 AD and 29 healthy controls) yields an RF accuracy of 81.60% and AUC of 0.8922, confirming the generalizability of the proposed approach. The results demonstrate that the proposed combination of adaptive decomposition, multi-domain feature extraction and optimization-driven feature selection provides an effective representation of EEG changes associated with cognitive decline. The methodology is computationally efficient and suitable for extension toward automated decision support tools for early detection of Alzheimer’s disease.
- 12 November 2026
- Published by World Scientific Pub Co Pte Ltd in Journal of Circuits, Systems and Computers
Automated Detection of Alzheimer’s Disease from EEG Signals Using VMD-Driven Multi-Domain Features and Rich and Poor Optimization-Enhanced Classification
Alzheimer’s disease is a progressive neurodegenerative disorder that affects large populations worldwide and produces measurable alterations in brain electrical activity. In this study, we propose an automated EEG-based methodology for distinguishing Alzheimer’s disease, mild cognitive impairment (MCI) and healthy control subjects. The EEG signals are first decomposed using adaptive variational mode decomposition (VMD) to obtain narrow-band intrinsic modes that capture key nonlinear and nonstationary properties of the recordings. A comprehensive set of time-domain, spectral, Hilbert envelope, and permutation entropy features is then extracted from each mode, resulting in an initial feature pool of 56 descriptors. These features are reduced using the rich and poor optimization (RPO) algorithm, and their discriminative ability is assessed using the Kruskal-Wallis statistical test with Bonferroni correction applied to control the family-wise error rate across all 56 features (corrected threshold [Formula: see text]). Forty-five statistically significant features are retained after Bonferroni correction and used as the final input feature set for classification. Three classical classifiers, namely random forest (RF), K-nearest neighbor (KNN) and support vector machine (SVM), are trained using 10-fold cross-validation with the synthetic minority oversampling technique (SMOTE) applied strictly within each training fold to prevent data leakage in order to evaluate the proposed framework. The RF model achieves the highest performance with a classification accuracy of 82.10%, a balanced accuracy of 71.90%, a macro-F1-score of 0.6534 and a macro area under the ROC curve (AUC) of 0.8933, supported by high sensitivity, specificity and predictive values. Cross-dataset validation on an independent publicly available EEG dataset (OpenNeuro ds004504, 36 AD and 29 healthy controls) yields an RF accuracy of 81.60% and AUC of 0.8922, confirming the generalizability of the proposed approach. The results demonstrate that the proposed combination of adaptive decomposition, multi-domain feature extraction and optimization-driven feature selection provides an effective representation of EEG changes associated with cognitive decline. The methodology is computationally efficient and suitable for extension toward automated decision support tools for early detection of Alzheimer’s disease.
- Open Access
- 30 July 2026
- Published by Springer Nature in Discover Applied Sciences
- Vol. 8 (11)
Explainable attention-based intrusion detection for encrypted 5G network traffic
The widespread adoption of end-to-end encryption in 5G networks limits the effectiveness of traditional intrusion detection systems that rely on payload inspection. This challenge is particularly critical for detecting Advanced Persistent Threats (APTs), which employ low-rate, long-duration, and stealthy communication strategies to evade conventional defenses. This study presents a privacy-preserving intrusion detection framework that operates exclusively on flow-level traffic metadata without deep packet inspection. Network packets are aggregated into bidirectional flows, from which temporal, statistical, and directional features are extracted to characterize behavioral patterns. A Transformer-based model with self-attention is employed to capture long-range dependencies across encrypted traffic sequences and identify subtle, temporally dispersed attack indicators. The framework is evaluated on a large-scale 5G-relevant dataset containing over one million flow records and compared against classical machine learning, ensemble, CNN, and LSTM models. Results demonstrate high recall and strong F1-score in distinguishing APT from benign traffic. Attention-based and feature-level explanations further reveal that prolonged communication, irregular timing gaps, and directional asymmetry significantly influence detection decisions. The findings support the practicality of explainable Transformer models for secure and scalable APT detection in encrypted 5G environments.
- 30 July 2026
- Published by Springer Nature in Discover Applied Sciences
- Vol. 8 (11)
Open AccessExplainable attention-based intrusion detection for encrypted 5G network traffic
The widespread adoption of end-to-end encryption in 5G networks limits the effectiveness of traditional intrusion detection systems that rely on payload inspection. This challenge is particularly critical for detecting Advanced Persistent Threats (APTs), which employ low-rate, long-duration, and stealthy communication strategies to evade conventional defenses. This study presents a privacy-preserving intrusion detection framework that operates exclusively on flow-level traffic metadata without deep packet inspection. Network packets are aggregated into bidirectional flows, from which temporal, statistical, and directional features are extracted to characterize behavioral patterns. A Transformer-based model with self-attention is employed to capture long-range dependencies across encrypted traffic sequences and identify subtle, temporally dispersed attack indicators. The framework is evaluated on a large-scale 5G-relevant dataset containing over one million flow records and compared against classical machine learning, ensemble, CNN, and LSTM models. Results demonstrate high recall and strong F1-score in distinguishing APT from benign traffic. Attention-based and feature-level explanations further reveal that prolonged communication, irregular timing gaps, and directional asymmetry significantly influence detection decisions. The findings support the practicality of explainable Transformer models for secure and scalable APT detection in encrypted 5G environments.
- 2 July 2026
- Published by Springer Nature
- p. 244-257
Smart Defenses: Leveraging Machine Learning to Strengthen IoT Network Traffic Security
- 2 July 2026
- Published by Springer Nature
- p. 244-257
Smart Defenses: Leveraging Machine Learning to Strengthen IoT Network Traffic Security
- 28 May 2026
- Published by Springer Nature in Neural Computing and Applications
- Vol. 38 (11)
Linking neural filters to cognitive states: a framework for EEG cluster analysis using neural networks
- 28 May 2026
- Published by Springer Nature in Neural Computing and Applications
- Vol. 38 (11)
Linking neural filters to cognitive states: a framework for EEG cluster analysis using neural networks
- 6 May 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Communications Standards Magazine
- Vol. PP (24712825) , 1-6
NextGen Smart-Ambulance via 6G: Semantic RAN Slicing and Edge Intelligence for Critical Care Mobility
The convergence of healthcare and next-generation wireless networks is reshaping digital medicine toward real-time, context-aware teleconsultations and remote interventions. However, existing 5G-based architectures face limitations in ultra-reliable low-latency communication (URLLC), network slicing autonomy, and context preservation across distributed healthcare nodes. This article presents a novel 6G-driven edge–cloud telemedicine framework that integrates an intelligent 6G stack and adaptive radio access network (RAN) orchestration layer to enable continuous, low-latency medical data exchange. The proposed design leverages AI-assisted RAN slicing, semantic communication channels, and blockchain-secured data pipelines for trustworthy healthcare delivery. A detailed case study on smart-ambulance system and how the 6G stack dynamically allocates edge resources, ensuring sub-millisecond response for diagnostic video streams and patient vitals. The experiential implementation on a prototype edge–cloud testbed validates end-to-end latency reduction by 37 % and reliability improvements exceeding 99.99 %. The results highlight the potential of 6G RAN–aware healthcare stacks to meet the stringent QoS and security demands of next-generation telemedicine. The study also discusses standardization implications for 3GPP, ETSI, and IEEE 6G working groups, emphasizing interoperability, privacy, and orchestration guidelines necessary for scalable digital health ecosystems.
- 6 May 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Communications Standards Magazine
- Vol. PP (24712825) , 1-6
NextGen Smart-Ambulance via 6G: Semantic RAN Slicing and Edge Intelligence for Critical Care Mobility
The convergence of healthcare and next-generation wireless networks is reshaping digital medicine toward real-time, context-aware teleconsultations and remote interventions. However, existing 5G-based architectures face limitations in ultra-reliable low-latency communication (URLLC), network slicing autonomy, and context preservation across distributed healthcare nodes. This article presents a novel 6G-driven edge–cloud telemedicine framework that integrates an intelligent 6G stack and adaptive radio access network (RAN) orchestration layer to enable continuous, low-latency medical data exchange. The proposed design leverages AI-assisted RAN slicing, semantic communication channels, and blockchain-secured data pipelines for trustworthy healthcare delivery. A detailed case study on smart-ambulance system and how the 6G stack dynamically allocates edge resources, ensuring sub-millisecond response for diagnostic video streams and patient vitals. The experiential implementation on a prototype edge–cloud testbed validates end-to-end latency reduction by 37 % and reliability improvements exceeding 99.99 %. The results highlight the potential of 6G RAN–aware healthcare stacks to meet the stringent QoS and security demands of next-generation telemedicine. The study also discusses standardization implications for 3GPP, ETSI, and IEEE 6G working groups, emphasizing interoperability, privacy, and orchestration guidelines necessary for scalable digital health ecosystems.
- 2 January 2026
- Published by Springer Nature
- p. 383-393
Layered Blockchain Resource Scheduling (LBRS) for Effective Blockchain Implementation in the Internet of Vehicles (IoV)
- 2 January 2026
- Published by Springer Nature
- p. 383-393
Layered Blockchain Resource Scheduling (LBRS) for Effective Blockchain Implementation in the Internet of Vehicles (IoV)
- Open Access
- 1 November 2025
- Published by Institute of Advanced Engineering and Science in Indonesian Journal of Electrical Engineering and Computer Science
- Vol. 40 (2) , 1087-1099
Interpretable federated deep learning models for predicting gait dynamics in biomechanics
Accurate prediction of human joint angle dynamics and reliable gait classifica tion are essential for applications in rehabilitation, biomechanics, and clinical monitoring. Traditional machine learning (ML) models trained on centralized data raise concerns about privacy, scalability, and transparency. This study proposes a federated deep learning (DL) framework that integrates privacy preserving model training with interpretable predictions. Specifically, a gated recurrent unit- deep neural network (GRU-DNN) hybrid model is developed for regression of joint angles, while a Long short-term memory- convolutional neural network (LSTM-CNN) hybrid model is designed for binary and multi class gait classification. The framework is deployed using the federated av eraging (FedAvg) algorithm across simulated clients, with each client training locally on its data. To enhance interpretability, the local interpretable model agnostic explanations (LIME) algorithm is integrated at the client level to gener ate human-understandable explanations for model predictions. The experimen tal results demonstrate significant improvements, including a reduction in global mean squared error (GMSE) from 56.16 to 3.31 and an increase in R-squared score from 0.80 to 0.99 for regression, along with classification accuracies of 0.97 (binary) and 0.94 (multi-class). This scalable, privacy-preserving frame work bridges the gap between accuracy and transparency, offering impactful applications in biomechanics, healthcare, and personalized medicine.
- 1 November 2025
- Published by Institute of Advanced Engineering and Science in Indonesian Journal of Electrical Engineering and Computer Science
- Vol. 40 (2) , 1087-1099
Open AccessInterpretable federated deep learning models for predicting gait dynamics in biomechanics
Accurate prediction of human joint angle dynamics and reliable gait classifica tion are essential for applications in rehabilitation, biomechanics, and clinical monitoring. Traditional machine learning (ML) models trained on centralized data raise concerns about privacy, scalability, and transparency. This study proposes a federated deep learning (DL) framework that integrates privacy preserving model training with interpretable predictions. Specifically, a gated recurrent unit- deep neural network (GRU-DNN) hybrid model is developed for regression of joint angles, while a Long short-term memory- convolutional neural network (LSTM-CNN) hybrid model is designed for binary and multi class gait classification. The framework is deployed using the federated av eraging (FedAvg) algorithm across simulated clients, with each client training locally on its data. To enhance interpretability, the local interpretable model agnostic explanations (LIME) algorithm is integrated at the client level to gener ate human-understandable explanations for model predictions. The experimen tal results demonstrate significant improvements, including a reduction in global mean squared error (GMSE) from 56.16 to 3.31 and an increase in R-squared score from 0.80 to 0.99 for regression, along with classification accuracies of 0.97 (binary) and 0.94 (multi-class). This scalable, privacy-preserving frame work bridges the gap between accuracy and transparency, offering impactful applications in biomechanics, healthcare, and personalized medicine.
- 28 October 2025
- Published by Springer Nature
- p. 422-447
Enhancing Breast Cancer Detection Through Explainable Federated Learning Models
- 28 October 2025
- Published by Springer Nature
- p. 422-447
Enhancing Breast Cancer Detection Through Explainable Federated Learning Models
- 14 October 2025
- Published by Springer Nature
- p. 353-380
Machine Learning for ESG Analytics: Clustering and Time-Series Forecasting for Sustainability Assessment
- 14 October 2025
- Published by Springer Nature
- p. 353-380
Machine Learning for ESG Analytics: Clustering and Time-Series Forecasting for Sustainability Assessment