Sripath Mishra

Sripath Mishra

CS PhD candidate at Purdue. Machine learning for UAV network security: intrusion detection with federated learning, RAG, and GNNs. Seeking Applied Scientist / ML Engineer roles.

About

I build machine learning systems that detect intrusions in dynamic UAV (drone) networks, where models have to stay accurate as the network topology, traffic, and threats shift in real time. My work combines federated learning, retrieval-augmented generation (RAG), and graph neural networks to make intrusion detection scalable across a swarm, generalizable to unseen attacks, and adaptive without constant retraining. I am a Computer Science PhD candidate at Purdue University, and I am pursuing Applied Scientist, Machine Learning Engineer, and Research Engineer roles with a 2027 start.

Most recently I developed FlowRAG, a retrieval-augmented intrusion detection approach for UAV networks (accepted at IEEE EDGE 2026), and released UAV-CAS, a calibrated digital-twin dataset and benchmark for intrusion detection in UAV swarm networks. Across projects I work end to end: simulation and data generation, model design and training, and evaluation against realistic adversarial conditions.

Research Interests

  • Explainable AI
  • Machine learning for security and intrusion detection
  • Federated and distributed learning
  • Retrieval-augmented generation (RAG) and LLMs for systems
  • Graph neural networks
  • Network traffic analysis and UAV / FANET security

Current Research

Intrusion Detection for UAV Networks (Jan 2025 – Present) · Advisor: Dr. Bharat Bhargava
Building scalable, generalizable, and adaptive intrusion detection for dynamic UAV networks using machine learning, retrieval-augmented methods, and federated learning. Outputs include the FlowRAG detection approach (IEEE EDGE 2026) and the UAV-CAS digital-twin dataset and benchmark.

Previous Research Experience

Network Traffic Classification (Jan 2023 – Jan 2025) · Advisor: Dr. Sonia Fahmy
Improved network traffic classification using topological data analysis, and built a large-scale network traffic generator that exposes the limits of current traffic classifiers. (IEEE IPCCC 2025.)

Multi-stream Network Analytics (Jan 2022 – Jan 2023)
Built multi-stream network analytics on a near-switch architecture to increase accuracy and reduce the reaction time of network telemetry systems.

Graph Neural Networks for Bug Prediction (Mar 2021 – Dec 2021)
Developed a GNN model that predicts bug locations in Python GitHub projects by combining code AST structure with text from GitHub issues.

Camera Network Analysis (Jan 2018 – May 2020)
Led a team building tools for automated discovery of real-time network camera feeds from heterogeneous web pages, plus a feed health checker, image database, and a camera-database API. (IEEE COMPSAC 2020; arXiv 2021.)

Education

Ph.D. in Computer Science (In Progress, expected 2027)
Purdue University, West Lafayette, IN — Advisor: Dr. Bharat Bhargava

M.S. in Computer Science (December 2021)
University of California, Los Angeles, CA

B.S. in Computer Science (May 2020)
Purdue University, West Lafayette, IN — Honors; Dean’s Honor List; Best Sophomore Award in Computer Science