On classification of multiple observations with application to marketing analysis

T. P. Logan, A. K. Gupta, Jie Chen

Research output: Contribution to journalArticle


In this paper, we study the likelihood procedure test (LPT) criterion for classification of the two-group multiple observations model, and derive the criterion under different parametric assumptions. A real life marketing data analysis is presented for illustrating the classification method introduced in this paper. It turns out that the LPT method is a useful alternative method in classification of multiple observations model.

Original languageEnglish (US)
Pages (from-to)387-402
Number of pages16
JournalAmerican Journal of Mathematical and Management Sciences
Issue number3-4
Publication statusPublished - Jan 1 2000
Externally publishedYes



  • Hierarchical model
  • Likelihood procedure
  • Multivariate normal distribution
  • Null distribution

ASJC Scopus subject areas

  • Business, Management and Accounting(all)
  • Applied Mathematics

Cite this