AIC Seminar Series

In this talk, I will describe our research on incorporating background knowledge in the form of feature hierarchies during Bayesian network learning. Feature hierarchies enable the learning system to aggregate categorical variables in meaningful ways, thus enabling an appropriate "discretization" for a categorical variable. In addition, by choosing the appropriate level of abstraction for the parent of a ndoe, we also support compact representations for the local probability models in the network. Our approach, Abstraction-Based Search (ABS), uses the feature hierarchies to search for an appropriate level of abstraction at which to represent each node in the network. (In fact, when a node appears as the parent of more than one descendant, it can be represented at a different level of abstraction for each of these different contexts.) I will describe three different heuristics that we have developed for guiding this search. We have applied ABS to several real-world domains. Our initial results support our hypothesis that feature hierarchies enable the learning system to identify networks that have better generalization performance. The resulting networks are more compact, require fewer parameters, and appear to capture the structure of the data more effectively. I will conclude with a number of interesting directions for future research in this area. This is joint work with Lise Getoor (UMd College Park) and Daphne Koller (Stanford University)

Dr. Marie desJardins is an assistant professor in the Department of Computer Science and Engineering at the University of Maryland, Baltimore County. Prior to joining the faculty in 2001, she was a senior computer scientist in the AI Center at SRI International in Menlo Park, California. Her research is in artificial intelligence, focusing on the areas of machine learning, planning, multi-agent systems, information management, reasoning with uncertainty, and decision theory.

Please arrive at least 10 minutes early in order to sign in and be escorted to the conference room. SRI is located at 333 Ravenswood Avenue in Menlo Park. Visitors may park in the visitors lot in front of Building E, and should follow the instructions by the lobby phone to be escorted to the meeting room. Detailed directions to SRI, as well as maps, are available from the Visiting AIC web page.

©2013 SRI International 333 Ravenswood Avenue, Menlo Park, CA 94025-3493
SRI International is an independent, nonprofit corporation. Privacy policy