anomal

CANShield: Deep Learning-Based Intrusion Detection Framework for Controller Area Networks at the Signal-Level

Fraud detection system using machine learning and deep learning (XGBoost + Autoencoder). Trains on synthetic financial transactions to flag suspicious activity with business-focused evaluation metrics.

University Assignment, it includes 3 different projects, about anomaly detection, kernel regression and stochastic text generation

Unsupervised anomaly detection for network traffic using Isolation Forest. Captures live packets, extracts statistical features, and flags threats in real-time via a React dashboard.

This repository hosts a comprehensive, web-based Anomaly Detection platform designed to identify irregularities across three distinct domains: Network Security, Industrial Manufacturing, and Medical Imaging. Built with Flask REST API and powered by PyTorch and Scikit-learn.

End-to-end real-time anomaly detection system leveraging Kafka streaming and stateful behavioural features for low-latency anomaly detection.