PATTERN CLUSTERING BY MULTIVARIATE MIXTURE ANALYSIS

Abstract

Cluster analysis is reformulated as a problem of estimating the parameters of a mixture of multivariate distributions. The maximum-likelihood theory and numerical solution techniques are developed for a fairly general class of distrubitions. The theory is applied to mixtures of multivariate normals ('NORMIX') and mixtures of multivariate Bernoulli distributions ('Latent Classes'). The feasibility of the procedures is demonstrated by two examples of computer solutions for normal mixture models of the Fisher Iris data and of artificially generated clusters with unequal covariance matrices.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Mar 01, 1969
Accession Number
AD0684087

Entities

People

  • John H. Wolfe

Tags

Communities of Interest

  • Human Systems

DTIC Thesaurus Topics

  • Artificial Intelligence
  • Bernoulli Distribution
  • Classification
  • Computational Science
  • Computer Programs
  • Computers
  • Data Science
  • Discrete Distribution
  • Estimators
  • Information Retrieval
  • Information Science
  • Naval Personnel
  • Navy
  • Normal Distribution
  • Pattern Recognition
  • Probability
  • Probability Distributions

Fields of Study

  • Mathematics

Readers

  • Combustion science or combustion engineering.
  • Operations Research
  • Regression Analysis.