## Exact Bayesian Pairwise Preference Learning and Inference on the Uniform Convex Polytope

author: Scott Sanner, NICTA, Australia's ICT Research Centre of Excellence
published: Jan. 24, 2012,   recorded: December 2011,   views: 63
Categories
You might be experiencing some problems with Your Video player.

# Slides

0:00 Slides Exact Bayesian Pairwise Preference Learning and Inference on the Uniform Convex Polytope Pic Part I: Bayesian Pairwise Preference Learning (BPPL) Learning from Preferences Bayesian Decision Theory Preference Feedback Types Preference Learning Feedback Part II: Bayesian Pairwise Preference Learning (BPPL) Bayesian Utility Model Response Likelihood Model Geometric View Uniform Prior Inference Tasks Part II: Bayesian Pairwise Preference Learning (BPPL) Polynomial Case Representation Polynomial Case Operations: - 1 Polynomial Case Operations: - 2 Polynomial Case Operations: max - 1 Polynomial Case Operations: max - 2 BPPL Inference Marginalization: ∫ - 1 Marginalization: ∫ - 2 Definite Integral Evaluation That’s It! Case ® XADD Maintaining Compact Cases Compactness of (X)ADDs Binary Operations on (X)ADDs XADD Maximization Maintaining XADD Orderings Correcting XADD Ordering XADD Pruning Empirical Results Future Work – Exact BPPL Thank you! Inference Tasks

# Report a problem or upload files

If you have found a problem with this lecture or would like to send us extra material, articles, exercises, etc., please use our ticket system to describe your request and upload the data.
Enter your e-mail into the 'Cc' field, and we will keep you updated with your request's status.

# Description

In Bayesian approaches to utility learning from preferences, the objective is to infer a posterior belief distribution over an agent’s utility function based on previously observed agent preferences. From this, one can then estimate quantities such as the expected utility of a decision or the probability of an unobserved preference, which can then be used to make or suggest future decisions on behalf of the agent. However, there remains an open question as to how one can represent beliefs over agent utilities, perform Bayesian updating based on observed agent pairwise preferences, and make inferences with this posterior distribution in an exact, closed-form. In this paper, we build on Bayesian pairwise preference learning models under the assumptions of linearly additive multi-attribute utility functions and a bounded uniform utility prior. These assumptions lead to a posterior form that is a uniform distribution over a convex polytope for which we then demonstrate how to perform exact, closed-form inference w.r.t. this posterior, i.e., without resorting to sampling or other approximation methods.