摘要
本文介绍了加权K-NN分类器的基本原理,并给出了相应的算法,着重对权函数的形式进行了讨论,提出了两种新的权函数定义公式.将加权K-NN分类器用在白血球自动分类系统中,分类精度比K-NN分类器有明显提高.
This paper introduces the basic principle of a weighted-K-NN classifier and presents the corresponding algorithm. The form of the weight function are discussed emphatically, and two new definition formulas of weighted function are produced. The weighted-K-NN Classifier has been applied to a system for the automated classification of white blood cells, and the classification accuracy is much higher than the K-NN Classifier.
出处
《自动化学报》
EI
CSCD
北大核心
1989年第2期174-177,共4页
Acta Automatica Sinica
关键词
KNN分类器
权函数
二分树
K-NN classifier
weighted function
weighted-K-NN classifier
white blood cells
binary tree