Comprehensive MATLAB Implementation of Hidden Markov Models (HMM)

This repository presents a comprehensive MATLAB-based toolkit for Hidden Markov Models (HMMs) developed by Cambridge University. The collection implements robust algorithms for various HMM operations with clear code structure and detailed documentation. Key implementations include: - HMM training algorithms (Baum-Welch/EM algorithm implementation with optimization techniques) - Hidden state inference methods (Forward-Backward algorithm with logarithmic scaling for numerical stability) - Model selection techniques (Bayesian Information Criterion implementation for optimal state selection) - Sequence generation functions (State transition and emission probability sampling with seed control) - Sequence alignment algorithms (Viterbi path optimization with dynamic programming implementation) The codebase features modular MATLAB functions with proper error handling and parameter validation. Each algorithm includes demonstration scripts showing practical applications with sample datasets. This collection serves as both an educational resource for understanding HMM mathematical foundations and a practical toolkit for research implementations, featuring well-commented code that illustrates probability calculations, expectation-maximization processes, and dynamic programming optimizations specific to HMM workflows.