Journal article
Error Rates of a Robust Classification Procedure Based on Dichotomous and Continuous Random Variables
Abstract
The error rates of the classification procedure based on Tiku's (1967, 1988) MML estimators are evaluated and shown to be smaller than those of the Chang and Afifi (1974) procedure which is based on sample means and sample variances.
Authors
Tuku ML; Balakrishnan N; Ambagaspittya RS
Journal
Communications in Statistics - Simulation and Computation, Vol. 18, No. 2, pp. 571–588
Publisher
Taylor & Francis
Publication Date
January 1989
DOI
10.1080/03610918908812777
ISSN
0361-0918