ESTIMATION IN A MIXTURE NORMAL DISTRIBUTION

  • Jee-Seon Baik (Department of Statistics pusan National university)
  • 발행 : 1997.03.01

초록

By Stochastic simulations we discuss the fitness of a mix-ture normal distribution to observations from general mixture distribu-tions using the MLE method and the EM algorithm. We calulate the probability of misclassifying objects and estimate the optimal number of mixture components with mutual information measure.

키워드

참고문헌

  1. J. Royal Statist. Soc. Ser. B (methodological) v.39 Maximum likelihood from incomplete data via the EM algorithm A. P. Dempster;N.M. Laird;D.B. Rubin
  2. Hidden Markov Models for Speech Recognition X.D. Huang;Y. Ariki;M.A. Jack