BayesianNetworks-MessagePassing-with-pyAgrum

Message Propagation in Bayesian Networks. A practice project analyzing message passing algorithms in both singly and multiply connected BNs using Python and pyAgrum.

🧠 Message Propagation in Bayesian Networks (ReseauxBayesiens)

This repository contains a practice project for a Graphical Models course, focusing on the core principles of message propagation used to perform inference in Bayesian Networks.

Key Topics Covered

The notebook explores the fundamental difference in inference techniques required for various network structures:

1. Singly Connected Networks (Polytrees): Demonstrates the efficiency of the belief propagation, where messages only need to be passed once without cycles.
2. Multiply Connected Networks (Graphs with Cycles): Illustrates the necessity of more complex algorithms, typically the **Junction Tree Algorithm **, to handle cycles and ensure correct inference.

🛠️ Technology and Libraries

• Python
• Jupyter Notebook
• pyAgrum: A powerful Python library for Bayesian Networks and related algorithms (used for modeling and inference).