Mahaveer Prasad
About
M.Tech student in Computer Science and Engineering at Delhi Technological University working across adversarial machine learning, backdoor and data-poisoning defense, robust network-intrusion detection and Android malware analysis. His public record includes four IEEE conference papers and a published Indian patent application for identifying and correcting poisoned training samples in deep neural networks.
Research focus · Adversarial Machine Learning, Backdoor and Data-Poisoning Defense, Network Intrusion Detection, Android Malware Analysis
Key Projects
ShortcutProbe research
the published study investigates inference-time backdoor-sample detection through internal stability analysis.
Multi-Signal Backdoor Defense research
the published framework combines model signals to detect poisoned samples and correct their labels.
Protocol-Aware Robustness Evaluation research
the published study evaluates deep-learning network intrusion-detection systems while preserving protocol semantics.
Android Malware Characterization research
the published study classifies Android malware using high-level behavioral features.
Patents
- Mahaveer Prasad, Vinod Kumar, Altaf Hussain and Rajeev Kumar, "Method and System for Identification and Correction of Poisoned Training Samples in Deep Neural Network," Indian Patent Application No. 202611018498, published April 10, 2026.
Publications · 4
- Conference paperA Unified Multi-Signal Indicator Framework for Backdoor Detection and Label Correction in Deep Neural NetworksMahaveer Prasad, Vinod Kumar2026 Fourth International Conference on Secure Cyber Computing and Communications (ICSCCC), pp. 368–373 · 2026DOI: 10.1109/ICSCCC69031.2026.11600051 ↗
- Conference paperShortcutProbe: Identifying Backdoor Samples Via Internal Stability AnalysisMahaveer Prasad, Ajit Kumar Yadav, Rajeev Kumar, Victor Fedoseev2026 14th International Symposium on Digital Forensics and Security (ISDFS), pp. 1–6 · 2026DOI: 10.1109/ISDFS69419.2026.11459068 ↗
- Conference paperProtocol-Aware Semantic Filtering for Reliable Adversarial Robustness Evaluation in DL-based Network Intrusion Detection SystemsMahaveer Prasad, Vinod Kumar2025 IEEE International Conference on Advanced Networks and Telecommunications Systems (ANTS), pp. 1–6 · 2025DOI: 10.1109/ANTS66931.2025.11430165 ↗
- Conference paperCharacterizing and Classifying Android Malware: A High-Level Feature ApproachMahaveer Prasad, Sanchit Gupta, Sarvjeet Kaur2023 International Conference on Quantum Technologies, Communications, Computing, Hardware and Embedded Systems Security (iQ-CCHESS), pp. 1–7 · 2023DOI: 10.1109/IQ-CCHESS56596.2023.10391357 ↗