LEARNING BAYESIAN NETWORKS: THE COMBINATION OF KNOWLEDGE AND STATISTICAL DATA

Yufeng Wu
Institute for Signal and Information Processing
Mississippi State University, Mississippi State, MS 39762
email: wu@isip.msstate.edu

ABSTRACT

A Bayesian network can be used to extract and encode knowledge from statistical data by using Bayesian statistical techniques. It has been shown to be remarkably effective when dealing with many data analysis problems. Two major advantages of using Bayesian network are its capabilities of handling incomplete data and learning causal relationship from a number of variables of interests. In this talk, we will discuss how to construct Bayesian networks from prior knowledge and how to apply Bayesian statistical methods on data sets in order to improve these models.

Additional items of interest: