Learning mixtures of ranking Models

Pranjal Awasthi, Avrim Blum, Or Sheffet, Aravindan Vijayaraghavan

Research output: Contribution to journalConference article

23 Citations (Scopus)

Abstract

This work concerns learning probabilistic models for ranking data in a heterogeneous population. The specific problem we study is learning the parameters of a Mallows Mixture Model. Despite being widely studied, current heuristics for this problem do not have theoretical guarantees and can get stuck in bad local optima. We present the first polynomial time algorithm which provably learns the parameters of a mixture of two Mallows models. A key component of our algorithm is a novel use of tensor decomposition techniques to learn the top-k prefix in both the rankings. Before this work, even the question of identifiability in the case of a mixture of two Mallows models was unresolved.

Original languageEnglish (US)
Pages (from-to)2609-2617
Number of pages9
JournalAdvances in Neural Information Processing Systems
Volume3
Issue numberJanuary
StatePublished - Jan 1 2014
Event28th Annual Conference on Neural Information Processing Systems 2014, NIPS 2014 - Montreal, Canada
Duration: Dec 8 2014Dec 13 2014

Fingerprint

Tensors
Polynomials
Decomposition
Statistical Models

All Science Journal Classification (ASJC) codes

  • Computer Networks and Communications
  • Information Systems
  • Signal Processing

Cite this

Awasthi, P., Blum, A., Sheffet, O., & Vijayaraghavan, A. (2014). Learning mixtures of ranking Models. Advances in Neural Information Processing Systems, 3(January), 2609-2617.
Awasthi, Pranjal ; Blum, Avrim ; Sheffet, Or ; Vijayaraghavan, Aravindan. / Learning mixtures of ranking Models. In: Advances in Neural Information Processing Systems. 2014 ; Vol. 3, No. January. pp. 2609-2617.
@article{6767d4ea5c2a4f899cb1b40d21184a42,
title = "Learning mixtures of ranking Models",
abstract = "This work concerns learning probabilistic models for ranking data in a heterogeneous population. The specific problem we study is learning the parameters of a Mallows Mixture Model. Despite being widely studied, current heuristics for this problem do not have theoretical guarantees and can get stuck in bad local optima. We present the first polynomial time algorithm which provably learns the parameters of a mixture of two Mallows models. A key component of our algorithm is a novel use of tensor decomposition techniques to learn the top-k prefix in both the rankings. Before this work, even the question of identifiability in the case of a mixture of two Mallows models was unresolved.",
author = "Pranjal Awasthi and Avrim Blum and Or Sheffet and Aravindan Vijayaraghavan",
year = "2014",
month = "1",
day = "1",
language = "English (US)",
volume = "3",
pages = "2609--2617",
journal = "Advances in Neural Information Processing Systems",
issn = "1049-5258",
number = "January",

}

Awasthi, P, Blum, A, Sheffet, O & Vijayaraghavan, A 2014, 'Learning mixtures of ranking Models', Advances in Neural Information Processing Systems, vol. 3, no. January, pp. 2609-2617.

Learning mixtures of ranking Models. / Awasthi, Pranjal; Blum, Avrim; Sheffet, Or; Vijayaraghavan, Aravindan.

In: Advances in Neural Information Processing Systems, Vol. 3, No. January, 01.01.2014, p. 2609-2617.

Research output: Contribution to journalConference article

TY - JOUR

T1 - Learning mixtures of ranking Models

AU - Awasthi, Pranjal

AU - Blum, Avrim

AU - Sheffet, Or

AU - Vijayaraghavan, Aravindan

PY - 2014/1/1

Y1 - 2014/1/1

N2 - This work concerns learning probabilistic models for ranking data in a heterogeneous population. The specific problem we study is learning the parameters of a Mallows Mixture Model. Despite being widely studied, current heuristics for this problem do not have theoretical guarantees and can get stuck in bad local optima. We present the first polynomial time algorithm which provably learns the parameters of a mixture of two Mallows models. A key component of our algorithm is a novel use of tensor decomposition techniques to learn the top-k prefix in both the rankings. Before this work, even the question of identifiability in the case of a mixture of two Mallows models was unresolved.

AB - This work concerns learning probabilistic models for ranking data in a heterogeneous population. The specific problem we study is learning the parameters of a Mallows Mixture Model. Despite being widely studied, current heuristics for this problem do not have theoretical guarantees and can get stuck in bad local optima. We present the first polynomial time algorithm which provably learns the parameters of a mixture of two Mallows models. A key component of our algorithm is a novel use of tensor decomposition techniques to learn the top-k prefix in both the rankings. Before this work, even the question of identifiability in the case of a mixture of two Mallows models was unresolved.

UR - http://www.scopus.com/inward/record.url?scp=84937883568&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84937883568&partnerID=8YFLogxK

M3 - Conference article

AN - SCOPUS:84937883568

VL - 3

SP - 2609

EP - 2617

JO - Advances in Neural Information Processing Systems

JF - Advances in Neural Information Processing Systems

SN - 1049-5258

IS - January

ER -

Awasthi P, Blum A, Sheffet O, Vijayaraghavan A. Learning mixtures of ranking Models. Advances in Neural Information Processing Systems. 2014 Jan 1;3(January):2609-2617.