Inference in Hidden Markov Models
Overview This repository contains a Jupyter notebook detailing work with Hidden Markov Models (HMMs) using the pgmpy library. The focus is on understanding, implementing, and experimenting with HMMs in various scenarios, including robot navigation and text improvement.
Objectives Learn and implement Hidden Markov Models from data. Use belief propagation for filtering and prediction queries. Implement the Viterbi algorithm to identify the most likely sequence of hidden states based on evidence. Apply HMMs to practical problems like robot navigation in a grid and improving misspelled text.
Key Concepts Understanding the structure and dynamics of HMMs. Developing proficiency in probabilistic queries including filtering, prediction, and evidence calculation.