RPCA Anomaly Detection with MATLAB Implementation

This document presents a comprehensive resource on RPCA (Robust Principal Component Analysis) anomaly detection. The project includes both custom-collected datasets and complete MATLAB implementation code. RPCA represents an advanced robust matrix decomposition technique that separates a given matrix into the sum of two components: a low-rank matrix and a sparse matrix. This method has widespread applications across signal processing, computer vision, and machine learning domains. The MATLAB implementation employs optimization algorithms such as the Augmented Lagrangian Multiplier method to solve the nuclear norm minimization problem. Key functions include: - Matrix preprocessing and normalization routines - Robust PCA decomposition using iterative thresholding - Anomaly scoring mechanisms based on sparse component analysis - Visualization tools for result interpretation This codebase serves as an educational and practical tool for researchers seeking to understand and apply RPCA anomaly detection techniques in their work, providing clear algorithmic implementations and ready-to-use examples.