CALCULATION OF BAYES RECOGNITION ERROR FOR 2 MULTIVARIATE GAUSSIAN DISTRIBUTIONS

被引：29

作者：

FUKUNAGA, K

KRILE, TF

机构：

[1] Department of Electrical Engineering, Purdue University, Lafayette

[2] Dennrtment of Electrical Engineering, Rose Polytechnic Institute, Terre Haute

来源：

IEEE TRANSACTIONS ON COMPUTERS | 1969年 / C 18卷 / 03期

关键词：

D O I：

10.1109/T-C.1969.222635

中图分类号：

TP3 [计算技术、计算机技术];

学科分类号：

0812 ;

摘要：

An algorithm is presented for calculating recognition error when applying pattern vectors to an optimum Bayes’ classifier. The pattern vectors are assumed to come from two classes whose populations have Gaussian statistics with unequal covariance matrices and arbitrary a priori probabilities. The quadratic discriminant function associated with a Bayes’ classifier is used as a one-dimensional random variable from which the probability of error is calculated, once the distribution of the discriminant function is obtained. A method for finding recognition error by approximating the distribution of the discriminant function is also given. This approximation, as well as the exact algorithm, is applied to an 8-dimensional example with good results. Copyright © 1969 by The Institute of Electrical and Electronics Engineers, Inc.

引用

页码：220 / &