摘要
本文提出对尤度判别(最大似然判别)指数表进行优化的一种简便方法——“双增计算法”。用该算法,指数表的判别符合率可随因素个数的增加而逐步提高,且最终可选到接近甚至达到最高判别符合率的因素组合及其指数表。本文还通过若干应用实例说明该算法具有简便易行、计算量小、效果好等优点。
A 'double-increase' algorithm for optimizing index number table of the Maximum Likelihood Discriminating Method (MLDM) is presented in this paper. By using the algorithm the discriminating comformable rate of the index number table can be stepwisely increased as the number k(1≤k≤m) of the factors building up the index number table increases and at the end we can easily select factor combination and its index number table with the highest (or approximately highest) diseriminating com- formable rate of 2~m-1 combinations of the m factors. The results of practical examples show that the algorithm is not only simple and feasible but also very efficient.
出处
《应用概率统计》
CSCD
北大核心
1991年第3期323-328,共6页
Chinese Journal of Applied Probability and Statistics
