Identifying Small Market Segments with Mixture Regression Models

The purpose of this work is to determine how well criteria designed to help the selection of the adequate number of market segments perform in recovering small niche market segments, in mixture regressions of normal data. As in real world data the true number of market segments is unknown, the results of this study are based on experimental data. The simulation experiment compares 27 segment retention criteria, comprising 14 information criteria and 13 classification-based criteria. The results reveal that AIC3, AIC4, HQ, BIC, CAIC, ICLBIC and ICOMPLBIC are the best criteria in recovering small niche segments and encourage its use.

Ana Oliveira-Brochado, Francisco Vitorino Martins (2014). Identifying Small Market Segments with Mixture Regression Models, International Journal of Finance, Insurance and Risk Management Volume 4 Issue 4, 812