摘要
KNN算法是应用最广泛的分类技术之一。文章简要介绍了KNN算法的基本原理,重点论述了研究人员针对KNN算法的不足所做的各种改进。主要从距离计算的改进、降低计算复杂度、K值的选择、与其它方法集成几个方面进行分析研究。
K-Nearest Neighbor Algorihtm is one classification technology,which have been used widely.This paper introduced the fundamental theory of KNN algorithm firstly,especially focused on many improvement methods which are researchers have been engageing in.The paper discussed mainly from aspects of distance calculation improvement,decreasing computational complexity,K-selection and integration of other methods with KNN algorithms.
出处
《科技信息》
2009年第31期81-81,383,共2页
Science & Technology Information
关键词
K-近邻算法
分类
距离加权
效率
K-Nearest Neighbor
Classification
Distance-weighted
Efficiency
