Bayesian network is applied widely in machine learning, data mining, diagnosis, etc; it has a solid evidence-based inference which is familiar to human intuition. However Bayesian network causes a little confusion because there are many complicated concepts, formulas and diagrams relating to it. Such concepts should be organized and presented in clear manner so as to be easy to understand it. This is the goal of this report. This report includes 4 main parts that cover principles of Bayesian network:
Part 1: Introduction to Bayesian network giving some basic concepts.
Part 2: Bayesian network inference discussing inference mechanism inside Bayesian network.
Part 3: Parameter learning tells us how to update parameters of Bayesian network.
Part 4: Structure learning surveys some main techniques to build up Bayesian network.

Keyword:Bayesian network, parameter learning, structure learning

• Full Text: PDF(457.71KB)
• Full Text: HTML