A mixture of mixture models for a classification problem: The unity measure error

Authors:
Marco Di Zio;Ugo Guarnera;Roberto Rocci
Affiliations:
Istituto Nazionale di Statistica, via Cesare Balbo 16, 00184 Roma, Italy;Istituto Nazionale di Statistica, via Cesare Balbo 16, 00184 Roma, Italy;Universití di Tor Vergata, via Columbia 2, 00133 Roma, Italy
Venue:
Computational Statistics & Data Analysis
Year:
2007

Citing 2
Cited 0

Identifiable finite mixtures of location models for clustering mixed-mode data

Statistics and Computing
Choosing starting values for the EM algorithm for getting the highest likelihood in multivariate Gaussian mixture models

Computational Statistics & Data Analysis

Quantified Score

Hi-index	0.03

Visualization

Abstract

A mixture of Gaussian mixture models is proposed to deal with the identification of survey respondents providing values in a wrong unity measure. The ''two-level'' mixture model allows effective classification in a non-normal setting. The natural constraints of the problem make the model identifiable. The effectiveness of the proposal is shown by simulation studies and an application to the 1997 Italian Labour Cost Survey.