Random Utility Theory for Social Choice

Abstract

Random utility theory models an agent's preferences on alternatives by drawing a real-valued score on each alternative (typically independently) from a parameterized distribution, and then ranking the alternatives according to scores. A special case that has received significant attention is the Plackett-Luce model, for which fast inference methods for maximum likelihood estimators are available. This paper develops conditions on general random utility models that enable fast inference within a Bayesian framework through MC-EM, providing concave loglikelihood functions and bounded sets of global maxima solutions. Results on both real-world and simulated data provide support for the scalability of the approach and capability for model selection among general random utility models including Plackett-Luce.

References in corpus (2)

• MCMC using Hamiltonian dynamics
• A Maximum Likelihood Approach For Selecting Sets of Alternatives

Cited by in corpus (14)

• Algorithmic Monoculture and Social Welfare
• Preference Elicitation For General Random Utility Models
• Minimax-optimal Inference from Partial Rankings
• A Statistical Framework of Watermarks for Large Language Models: Pivot, Detection Efficiency and Optimal Rules
• Learning Mixed Multinomial Logit Model from Ordinal Data
• Learning from Comparisons and Choices
• Strategy for Boosting Pair Comparison and Improving Quality Assessment Accuracy
• Choice by Elimination via Deep Neural Networks
• Fairness in Ranking under Uncertainty
• Regret Minimization in Stochastic Contextual Dueling Bandits
• A Bayesian Choice Model for Eliminating Feedback Loops
• Best-item Learning in Random Utility Models with Subset Choices
• Adversarial Dueling Bandits
• Dueling Bandits with Adversarial Sleeping