Vasudeva Pai
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Vasudeva Pai has a total of 7 co-authors. Here are the top co-authors by number of shared publications:
- B. H. Karthik Pai
Visvesvaraya Technological University, Belagavi, India
1 shared publication - Devidas
Nitte University (Deemed to be), NMAM Institute of Technology
1 shared publication - G. S. Sudhiksha
Department of Information Science and Engineering, Nitte (Deemed to be University), NMAM Institute of Technology (NMAMIT), Udupi, India
1 shared publication - K. Varsha
Department of Information Science and Engineering, Nitte (Deemed to be University), NMAM Institute of Technology (NMAMIT), Udupi, India
1 shared publication - Manjunatha S.
Manipal Institute of Technology Manipal, Manipal Academy of Higher Education, Udupi, India
1 shared publication - N. D. Adesh
Manipal Academy of Higher Education
1 shared publication
Latest Publications
- Open Access
- 29 July 2025
- Published by Springer Nature in International Journal of Computational Intelligence Systems
- Vol. 18 (1) , 1-17
Systematic Approach for Malware Detection in IoT Devices: Enhancing Security and Performance
The proliferation of Internet of Things (IoT) devices has introduced significant security challenges, particularly concerning the detection and mitigation of malware threats. This study presents a systematic approach to malware detection that aims to improve both the security and performance of IoT systems. Using the IoT23 dataset, which contains a wide range of network traffic patterns from various IoT devices and malware families, the research explores and evaluates multiple machine learning techniques. These include ensemble methods such as Bagging, Stacking, Voting, AdaBoost, and H2O AutoML, as well as advanced models such as sparse neural networks with pruning and feature selection and regularized classifiers L1. The primary objective is to develop lightweight yet highly accurate models suitable for deployment on resource-constrained IoT devices. A comprehensive comparison of these techniques demonstrates the importance of achieving a balance between detection accuracy and computational efficiency. Among the models evaluated, the SNIPE approach shows the best performance, achieving an accuracy of 91.9% while maintaining minimal computational overhead. This makes it particularly well suited for real-world IoT environments, where performance and energy efficiency are critical. The findings of this study provide valuable insights for the development of robust, scalable, and resource-aware malware detection systems, laying a strong foundation for future research and practical cybersecurity solutions in the rapidly evolving IoT landscape.
- 29 July 2025
- Published by Springer Nature in International Journal of Computational Intelligence Systems
- Vol. 18 (1) , 1-17
Open AccessSystematic Approach for Malware Detection in IoT Devices: Enhancing Security and Performance
The proliferation of Internet of Things (IoT) devices has introduced significant security challenges, particularly concerning the detection and mitigation of malware threats. This study presents a systematic approach to malware detection that aims to improve both the security and performance of IoT systems. Using the IoT23 dataset, which contains a wide range of network traffic patterns from various IoT devices and malware families, the research explores and evaluates multiple machine learning techniques. These include ensemble methods such as Bagging, Stacking, Voting, AdaBoost, and H2O AutoML, as well as advanced models such as sparse neural networks with pruning and feature selection and regularized classifiers L1. The primary objective is to develop lightweight yet highly accurate models suitable for deployment on resource-constrained IoT devices. A comprehensive comparison of these techniques demonstrates the importance of achieving a balance between detection accuracy and computational efficiency. Among the models evaluated, the SNIPE approach shows the best performance, achieving an accuracy of 91.9% while maintaining minimal computational overhead. This makes it particularly well suited for real-world IoT environments, where performance and energy efficiency are critical. The findings of this study provide valuable insights for the development of robust, scalable, and resource-aware malware detection systems, laying a strong foundation for future research and practical cybersecurity solutions in the rapidly evolving IoT landscape.
- Open Access
- 1 January 2021
- Published by IOP Publishing in IOP Conference Series: Materials Science and Engineering
- Vol. 1013 (1) , 012038
Comparative analysis of Machine Learning algorithms for Intrusion Detection
In this modern era, the network related applications, programs and services are growing enormously but the network security issues also grow along with them. Keeping the network secure is a challenging and a crucial task. To maintain the secure network there must be some system which can detect and identify any malicious activity happening in network. This system is called as Intrusion Detection System. There are many traditional network security tools and techniques of preventing intrusion like firewalls, anti-virus, encryption-decryption, access control etc. But all are not effective in protecting network from increasing attacks. The network traffic can be categories into normal and intrusive traffic using Machine Learning (ML) algorithms. Here, the preliminary comparative study regarding which type of machine learning algorithm performs better in identifying the attacks namely Denial of Service, Probe, User to Root and Remote to Local. The NSL-KDD dataset is used to study features and behavior of malicious attacker using machine learning techniques. This study can be taken as reference for mechanical engineers for developing a safe automation in industrial atmosphere and automation in automobile.
- 1 January 2021
- Published by IOP Publishing in IOP Conference Series: Materials Science and Engineering
- Vol. 1013 (1) , 012038
Open AccessComparative analysis of Machine Learning algorithms for Intrusion Detection
In this modern era, the network related applications, programs and services are growing enormously but the network security issues also grow along with them. Keeping the network secure is a challenging and a crucial task. To maintain the secure network there must be some system which can detect and identify any malicious activity happening in network. This system is called as Intrusion Detection System. There are many traditional network security tools and techniques of preventing intrusion like firewalls, anti-virus, encryption-decryption, access control etc. But all are not effective in protecting network from increasing attacks. The network traffic can be categories into normal and intrusive traffic using Machine Learning (ML) algorithms. Here, the preliminary comparative study regarding which type of machine learning algorithm performs better in identifying the attacks namely Denial of Service, Probe, User to Root and Remote to Local. The NSL-KDD dataset is used to study features and behavior of malicious attacker using machine learning techniques. This study can be taken as reference for mechanical engineers for developing a safe automation in industrial atmosphere and automation in automobile.