Collaborative ordinal regression

Published 1 January 2006

Shipeng Yu, Kai Yu, Volker Tresp, Hans‐Peter Kriegel

Citations54

TL;DR

Empirical studies show that the collaborative ordinal regression model outperforms the individual counterpart in preference learning applications and explores the dependency between ranking functions through a hierarchical Bayesian model and assign a common Gaussian Process prior to all individual functions.

Abstract

Ordinal regression has become an effective way of learning user preferences, but most research focuses on single regression problems. In this paper we introduce collaborative ordinal regression, where multiple ordinal regression tasks are handled simultaneously. Rather than modeling each task individually, we explore the dependency between ranking functions through a hierarchical Bayesian model and assign a common Gaussian Process (GP) prior to all individual functions. Empirical studies show that our collaborative model outperforms the individual counterpart in preference learning applications.

Keywords

Computer Science

The MIT Press eBooksGaussian Processes for Machine Learning

10,409 Citations2005Carl Edward Rasmussen, Christopher K. I. Williams

The book provides a long-needed, systematic and unified treatment of theoretical and practical aspects of GPs in machine learning, targeted at researchers and students in machine learning and applied statistics.

Journal of the Royal Statistical Society Series B (Statistical Methodology)Regression Models for Ordinal Data

4,395 Citations1980Peter McCullagh

Learning to rank using gradient descent

2,759 Citations2005Chris Burges, Tal Shaked +5 more

RankNet is introduced, an implementation of these ideas using a neural network to model the underlying ranking function, and test results on toy data and on data from a commercial internet search engine are presented.

ACM SIGIR ForumIR evaluation methods for retrieving highly relevant documents

1,466 Citations2017Kalervo Järvelin, Jaana Kekäläinen

The novel evaluation methods and the case demonstrate that non-dichotomous relevance assessments are applicable in IR experiments, may reveal interesting phenomena, and allow harder testing of IR methods.

Fast maximum margin matrix factorization for collaborative prediction

968 Citations2005Jasson D. M. Rennie, Nathan Srebro

This work investigates a direct gradient-based optimization method for MMMF and finds that MMMf substantially outperforms all nine methods he tested and demonstrates it on large collaborative prediction problems.

DSpace@MIT (Massachusetts Institute of Technology)A family of algorithms for approximate bayesian inference

902 Citations2001Thomas P. Minka, Rosalind W. Picard

This thesis presents an approximation technique that can perform Bayesian inference faster and more accurately than previously possible, and is found to be convincingly better than rival approximation techniques: Monte Carlo, Laplace's method, and variational Bayes.

View PDF

The MIT Press eBooksPranking with Ranking

550 Citations2002Koby Crammer, Yoram Singer

A simple and efficient online algorithm is described, its performance in the mistake bound model is analyzed, its correctness is proved, and it outperforms online algorithms for regression and classification applied to ranking.

Learning Gaussian processes from multiple tasks

388 Citations2005Kai Yu, Volker Tresp +1 more

This work considers the problem of multi-task learning, that is, learning multiple related functions, and presents a hierarchical Bayesian framework, that exploits the equivalence between parametric linear models and nonparametric Gaussian processes.

Scientific Repository (Petra Christian University)Gaussian Processes for Ordinal Regression

373 Citations2005Wei Chu, Zoubin Ghahramani

Preference learning with Gaussian processes

362 Citations2005Wei Chu, Zoubin Ghahramani

A probabilistic kernel approach to preference learning based on Gaussian processes and a new likelihood function is proposed to capture the preference relations in the Bayesian framework.