Please note: This PhD seminar will be given online.
Alister
Liao, PhD
candidate
David
R.
Cheriton
School
of
Computer
Science
Supervisor: Professor Peter van Beek
Bayesian networks (BNs) are widely used as a data analytics tool in diverse areas, including finance, medicine, and sports. Learning the structure of a BN from data can be cast as an optimization problem using the well-known score-and-search approach. However, selecting a single model (i.e., the best scoring BN) can be misleading or may not achieve the best possible accuracy. An alternative to committing to a single model is to perform some form of Bayesian or frequentist model averaging, where the space of possible BNs is sampled or enumerated in some fashion.
Unfortunately, existing approaches for model averaging either severely restrict the structure of the Bayesian network or have only been shown to scale to networks with fewer than 30 random variables. In this talk, I will discuss our recent progress in scaling up the structure learning algorithm for model averaging using different scores (model selection criteria). In particular, we propose a novel approach to collect near optimal BNs inspired by performance guarantees in approximation algorithms that scales to significantly larger BNs than previous approaches. Then, I will present our recent investigation of different scores under the model averaging framework. Aided by our new approach, we are able to address design limits in previous score comparison studies as we set to find the most promising scores for structure discovery and density estimation.
To join this PhD seminar on MS Teams, please go to https://teams.microsoft.com/l/meetup-join/19%3ameeting_ZWFlOWI0ZTUtOTg4Ny00OWZkLWE5ZTctN2FiZTA3NDY2M2E1%40thread.v2/0?context=%7b%22Tid%22%3a%22723a5a87-f39a-4a22-9247-3fc240c01396%22%2c%22Oid%22%3a%2255a4dd3f-4336-4fc1-920f-697bade427ea%22%7d.