ResearchTrend.AI
  • Papers
  • Communities
  • Events
  • Blog
  • Pricing
Papers
Communities
Social Events
Terms and Conditions
Pricing
Parameter LabParameter LabTwitterGitHubLinkedInBlueskyYoutube

© 2025 ResearchTrend.AI, All rights reserved.

  1. Home
  2. Papers
  3. 1502.04631
19
21

Clustering and Inference From Pairwise Comparisons

16 February 2015
Rui Wu
Jiaming Xu
R. Srikant
Laurent Massoulié
Marc Lelarge
Bruce Hajek
ArXivPDFHTML
Abstract

Given a set of pairwise comparisons, the classical ranking problem computes a single ranking that best represents the preferences of all users. In this paper, we study the problem of inferring individual preferences, arising in the context of making personalized recommendations. In particular, we assume that there are nnn users of rrr types; users of the same type provide similar pairwise comparisons for mmm items according to the Bradley-Terry model. We propose an efficient algorithm that accurately estimates the individual preferences for almost all users, if there are rmax⁡{m,n}log⁡mlog⁡2nr \max \{m, n\}\log m \log^2 nrmax{m,n}logmlog2n pairwise comparisons per type, which is near optimal in sample complexity when rrr only grows logarithmically with mmm or nnn. Our algorithm has three steps: first, for each user, compute the \emph{net-win} vector which is a projection of its (m2)\binom{m}{2}(2m​)-dimensional vector of pairwise comparisons onto an mmm-dimensional linear subspace; second, cluster the users based on the net-win vectors; third, estimate a single preference for each cluster separately. The net-win vectors are much less noisy than the high dimensional vectors of pairwise comparisons and clustering is more accurate after the projection as confirmed by numerical experiments. Moreover, we show that, when a cluster is only approximately correct, the maximum likelihood estimation for the Bradley-Terry model is still close to the true preference.

View on arXiv
Comments on this paper