Kapil Shrivastava
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Kapil Shrivastava has a total of 5 co-authors. Here are the top co-authors by number of shared publications:
- Munish Khanna
School of Computer Science and Engineering, Galgotias University, Greater Noida, India
2 shared publications - Law Kumar Singh
Amity University, Mohali, India
1 shared publication - Manish Tiwari
Sir Padampat Singhania University
1 shared publication - Prasun Chakrabarti
Sir Padampat Singhania University, Udaipur, Rajasthan, India
1 shared publication - Rekha Singh
Department of Physics, Uttar Pradesh Rajarshi Tandon Open University, Prayagraj, India
1 shared publication
Latest Publications
- 8 October 2025
- Published by Turkish Journal of Engineering in Turkish Journal of Engineering
- Vol. 9 (4) , 738-753
AtEdO-BLGBM: A Hybrid Optimization and Boosting Framework for Scalable Intrusion Detection in Heterogeneous IoT Networks
IIntrusion detection is a network security measure that regularly tracks the data flow and prevents malicious activities in Internet of Things (IoT) Networks. A significant rise in connected devices has led to an increase in threats across the IoT. Numerous traditional techniques have been introduced to achieve significant attack detection; however, they have resulted in certain limitations in terms of flexibility, scalability, and performance in heterogeneous infrastructures. To overcome these challenges, this research proposes an Attack-Evaded Optimization Ensemble Bi-directional Long Short-Term Memory Gradient-Boosting Machine (AtEdO-BLGBM) model, which effectively detects intrusions in complex and dynamic network environments. The AtEdO algorithm fine-tunes parameters with better convergence and improved global optima in intrusion detection. Moreover, the proposed model utilizes the AtEdO generative adversarial network to handle class labels, thereby balancing the imbalanced data distribution in heterogeneous networks. The combined BLGBM approach reduces computational costs and enhances the training process, thereby improving detection capability. Finally, the AtEdO-BLGBM model yields 96.78% accuracy, 96.71% negative predictive value and 97.03% positive predictive value, demonstrating its effectiveness in accurate and efficient intrusion detection. In future work, we aim to integrate transformer-based architectures and energy-efficient optimization techniques to enhance real-time detection.
- 8 October 2025
- Published by Turkish Journal of Engineering in Turkish Journal of Engineering
- Vol. 9 (4) , 738-753
AtEdO-BLGBM: A Hybrid Optimization and Boosting Framework for Scalable Intrusion Detection in Heterogeneous IoT Networks
IIntrusion detection is a network security measure that regularly tracks the data flow and prevents malicious activities in Internet of Things (IoT) Networks. A significant rise in connected devices has led to an increase in threats across the IoT. Numerous traditional techniques have been introduced to achieve significant attack detection; however, they have resulted in certain limitations in terms of flexibility, scalability, and performance in heterogeneous infrastructures. To overcome these challenges, this research proposes an Attack-Evaded Optimization Ensemble Bi-directional Long Short-Term Memory Gradient-Boosting Machine (AtEdO-BLGBM) model, which effectively detects intrusions in complex and dynamic network environments. The AtEdO algorithm fine-tunes parameters with better convergence and improved global optima in intrusion detection. Moreover, the proposed model utilizes the AtEdO generative adversarial network to handle class labels, thereby balancing the imbalanced data distribution in heterogeneous networks. The combined BLGBM approach reduces computational costs and enhances the training process, thereby improving detection capability. Finally, the AtEdO-BLGBM model yields 96.78% accuracy, 96.71% negative predictive value and 97.03% positive predictive value, demonstrating its effectiveness in accurate and efficient intrusion detection. In future work, we aim to integrate transformer-based architectures and energy-efficient optimization techniques to enhance real-time detection.
- 29 August 2025
- Published by Springer Nature in SN Computer Science
- Vol. 6 (7) , 1-14
CFA–ASA: An Adaptive Hybrid Feature Selection Approach for Thyroid Disease Prediction
- 29 August 2025
- Published by Springer Nature in SN Computer Science
- Vol. 6 (7) , 1-14
CFA–ASA: An Adaptive Hybrid Feature Selection Approach for Thyroid Disease Prediction