Estimation of Parameters for a Mixture of Normal Distributions
作者:
Victor Hasselblad,
期刊:
Technometrics
(Taylor Available online 1966)
卷期:
Volume 8,
issue 3
页码: 431-444
ISSN:0040-1706
年代: 1966
DOI:10.1080/00401706.1966.10490375
出版商: Taylor & Francis Group
数据来源: Taylor
摘要:
nobservations are taken from a mixture ofKnormal subpopulations, where the value ofKis known. It is assumed that thesenobservations are given asNfrequencies from equally spaced intervals. Initial guesses of theKmeans,Kvariances, andK− 1 proportions are made using the maximum likelihood estimates for a single truncated normal population as derived by Hald. Then an approximation to the likelihood function of the entire sample is used, and attempts to maximize this yield two iteration formulas. In practice, the method of steepest descent always converged, although the rate was not always fast. Special cases of equal variances and variances proportional to the square of the mean are also considered.
点击下载:
PDF (786KB)
返 回