A tight convex upper bound on the likelihood of a finite mixture

Elad Mezuman, Yair Weiss

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

The likelihood function of a finite mixture model is a non-convex function with multiple local maxima and commonly used iterative algorithms such as EM will converge to different solutions depending on initial conditions. In this paper we ask: is it possible to assess how far we are from the global maximum of the likelihood? Since the likelihood of a finite mixture model can grow unboundedly by centering a Gaussian on a single datapoint and shrinking the covariance, we constrain the problem by assuming that the parameters of the individual models are members of a large discrete set (e.g. estimating a mixture of two Gaussians where the means and variances of both Gaussians are members of a set of a million possible means and variances). For this setting we show that a simple upper bound on the likelihood can be computed using convex optimization and we analyze conditions under which the bound is guaranteed to be tight. This bound can then be used to assess the quality of solutions found by EM (where the final result is projected on the discrete set) or any other mixture estimation algorithm. For any dataset our method allows us to find a finite mixture model together with a dataset-specific bound on how far the likelihood of this mixture is from the global optimum of the likelihood.

Original languageEnglish
Title of host publication2016 23rd International Conference on Pattern Recognition, ICPR 2016
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages1683-1688
Number of pages6
ISBN (Electronic)9781509048472
DOIs
StatePublished - 1 Jan 2016
Event23rd International Conference on Pattern Recognition, ICPR 2016 - Cancun, Mexico
Duration: 4 Dec 20168 Dec 2016

Publication series

NameProceedings - International Conference on Pattern Recognition
Volume0

Conference

Conference23rd International Conference on Pattern Recognition, ICPR 2016
Country/TerritoryMexico
CityCancun
Period4/12/168/12/16

All Science Journal Classification (ASJC) codes

  • Computer Vision and Pattern Recognition

Fingerprint

Dive into the research topics of 'A tight convex upper bound on the likelihood of a finite mixture'. Together they form a unique fingerprint.

Cite this