Title: Uncertainty quantification in the Bradley-Terry-Luce model
Authors: Anderson Ye Zhang - University of Pennsylvania (United States) [presenting]
Abstract: Ranking from pairwise comparisons is a central problem in a wide range of learning and social contexts. The Bradley-Terry-Luce (BTL) model is one of the most studied models for analyzing ranking data. Despite all the recent progress, uncertainty quantification under the BTL model remains unclear. To address this challenge, we first establish non-asymptotic entrywise distributions of the maximum likelihood estimation and the spectral method under the BTL model. We then develop statistical inference procedures for individual rankings and preference parameters.