Timezone: »

Scaling MPE Inference for Constrained Continuous Markov Random Fields with Consensus Optimization
Stephen H Bach · Matthias Broecheler · Lise Getoor · Dianne P O'Leary

Tue Dec 04 07:00 PM -- 11:59 PM (PST) @ Harrah’s Special Events Center 2nd Floor #None

Probabilistic graphical models are powerful tools for analyzing constrained, continuous domains. However, finding most-probable explanations (MPEs) in these models can be computationally expensive. In this paper, we improve the scalability of MPE inference in a class of graphical models with piecewise-linear and piecewise-quadratic dependencies and linear constraints over continuous domains. We derive algorithms based on a consensus-optimization framework and demonstrate their superior performance over state of the art. We show empirically that in a large-scale voter-preference modeling problem our algorithms scale linearly in the number of dependencies and constraints.

Author Information

Stephen H Bach (University of Maryland)
Matthias Broecheler (University of Maryland CP)
Lise Getoor (UC Santa Cruz)

Lise Getoor is an Associate Professor in the Computer Science Department and the Institute for Advanced Computer Studies at the University of Maryland, College Park. Her research areas include machine learning, reasoning under uncertainty, and database management. She is co-editor with Ben Taskar of the book 'An Introduction to Statistical Relational Learning', MIT Press, 2007. She is a board member of the International Machine Learning Society, and has served as Machine Learning Journal Action Editor, Associate Editor for the ACM Transactions of Knowledge Discovery from Data, JAIR Associate Editor, and on the AAAI Council. She is a recipient of several best paper awards, an NSF Career Award and a National Physical Sciences Consortium Fellowship. She received her PhD from Stanford University, her Master’s degree from the University of California, Berkeley, and her undergraduate degree from the University of California, Santa Barbara.

Dianne P O'Leary (University of Maryland)

More from the Same Authors