论文部分内容阅读
本文给出图象识别中较之NN和k-NN分类规则更为一般的W-k-NN分类规则,指出这两个应用很广的规则仅为本文结果的特例。文中导出了一系列样本精简的W-k-NN算法,研究了W-k-NN分类规则的错误概率并实现了W-k-NN分类器的程序设计。本文的结果完善了文献[1]中给出的分类模型。
In this paper, W-k-NN classification rules which are more general than NN and k-NN classification rules are given. It is pointed out that these two widely used rules are only special cases of this paper. In this paper, a series of sample-reduced W-k-NN algorithms are derived. The W-k-NN classification rules are investigated for error probability and the W-k-NN classifier programming is implemented. The results in this paper improve the classification model given in [1].