Estimating the Number of Clusters in Logistic Regression Clustering by an Information Theoretic Criterion
This paper studies the problem of estimating the number of clusters in the context of logistic regression clustering. The classi.cation likelihood approach is employed to tackle this problem. An information theoretic criterion for selecting the number of logistic curves is proposed in the sequel and then its asymptotic property is considered.
The paper is arranged as follows: In Section 2, some notations are given and an information theoretic criterion is proposed for estimating the number of clusters. In Section 3, the small sample performance of the proposed criterion is studied by Monte Carlo simulation. In Section 4, the asymptotic property of the criterion proposed in Section 2 is investigated. Some lemmas needed in Section 4 are given in the appendix.
KeywordsLogistic Regression Binomial Distribution Maximum Likelihood Estimator Asymptotic Property Linear Predictor
Chung KL (2001) A Course in Probability Theory (3rd edition). Academic PressGoogle Scholar
Farewell BT, Sprott D (1988) The use of a mixture model in the analysis of count data. Biometrics 44:1191-1194MATHCrossRefGoogle Scholar
Follmann DA, Lambert D (1989) Generalizing logistic regression by nonparametric mixing. Journal of the American Statistical Association 84:295-300CrossRefGoogle Scholar
Follmann DA, Lambert D (1991) Identifiability for nonparametric mixtures of logistic regressions. Journal of Statistical Planning and Inference 27:375-381MATHCrossRefMathSciNetGoogle Scholar
McCullagh P, Nelder JA (1989) Generalized Linear Models (2nd edition). Chapman and HallGoogle Scholar
Qian G, Field C (2002) Law of iterated logarithm and consistent model selection criterion in logistic regression. Statistics & Probability Letters 56:101-112MATHCrossRefMathSciNetGoogle Scholar
Shao Q, Wu Y (2005) A consistent procedure for determining the number of clusters in regression clustering. Journal of Statistical Planning and Inference 135:461-476MATHCrossRefMathSciNetGoogle Scholar
Wedel M, DeSarbo WS (1995) A mixture likelihood approach for generalized linear models. Journal of Classification 12:21-55MATHCrossRefGoogle Scholar
© Physica-Verlag Heidelberg 2008