I research machine learning-based intrusion detection for IoT systems, with a focus on streaming evaluation, concept drift, conformal evaluation, adaptive retraining, and explainable AI.
I am a PhD student in Computer and Cyber Sciences at Augusta University. My work sits at the intersection of cybersecurity, machine learning, IoT systems, digital forensics, and real-time intrusion detection.
My current research focuses on building intrusion detection systems that can operate under realistic streaming conditions, where traffic distributions shift over time and models need to adapt without collapsing under retraining overhead.
I also teach introductory Python as a graduate teaching assistant and work part-time as a Senior Software Engineer at Kart Chaser, where I help maintain live-streaming and cloud infrastructure.
- Streaming IoT intrusion detection
- Conformal evaluation for drift detection
- Adaptive retraining and adaptive chunking
- Explainable AI for security systems
- Multimodal security data pipelines
- Cloud and live-streaming infrastructure
- Python tooling, automation, and research software
Cybersecurity Machine Learning IoT Security
Digital Forensics Concept Drift Conformal Evaluation
XAI Network Traffic Edge Feasibility
A research framework for real-time intrusion detection using conformal evaluation, adaptive retraining, and streaming simulation under concept drift.
Core ideas:
- Drift-aware model evaluation
- Approximate cross-conformal evaluation
- Adaptive chunk sizing
- Runtime-aware retraining policies
- Comparative analysis against representation-space drift detectors
A machine learning-based intrusion detection system for IoT network traffic, integrating real-time flow classification and explainable AI components.
Focus areas:
- Binary and multiclass attack detection
- CICFlowMeter-style feature pipelines
- SHAP and LIME-based explanations
- Reproducible Python package structure
Research and infrastructure tooling for cybersecurity experiments, GitHub Actions runners, paper workflows, and lab automation.
My research includes work on:
- IoT malware and attack behavior
- ML-based network anomaly detection
- BGP security policy analysis
- IoT device fingerprinting and authentication
- Privacy and security implications in consumer technologies
- Vault apps and gray-zone digital forensics
For a more complete list, visit my website: sethbarrett.xyz
- Research code that other people can actually reproduce
- Security experiments with realistic deployment assumptions
- Python CLIs and developer tools
- Cloud infrastructure for streaming systems
- Homelab setups for research and teaching
- Clean documentation for messy technical systems
- Website: sethbarrett.xyz
- GitHub: @sethbarrett50
- DFAIR Lab: DFAIR-LAB-Augusta
Researching adaptive, explainable, and deployable security systems for real-world IoT environments.




