摘要
k-近质心近邻原则是k-近邻原则的一种有效扩展,是有效的模式分类方法之一。k-近质心近邻原则容易受到局外点的影响;同时,所有的k-近质心近邻点在分类决策时具有相同的权重和分类贡献率,这显然是不合理的。为了解决这一问题,考虑到质心近邻在模式分类问题上具有近邻特性和空间分布特性,提出一种基于局部权重的近质心近邻算法,实验结果表明该LWKNCN算法在分类精度上优于传统的KNN算法和KNCN算法。
The k-nearest centroid neighbor rule （KNCN）, as an effective extension of the k-Nearest Neighbor rule（ KNN）, is one of the effective algorithms in pattern classification. The KNCN is prone to be seriously influenced bythe existing outliers. At the same time, all the k-nearest centroid neighbor samples have the same weight and thesame contribution to classification results, which is unreasonable. To solve this problem, this paper proposes a nea-rest centroid neighbor algorithm based on the local weight, taking account of the proximity and spatial distributioncharacteristics of the neighbors for a query pattern. The experimental results show that the classification accuracy ofLWKNCN is better than that of the traditional KNN algorithm and KNCN algorithm.
出处
《应用科技》
CAS
2015年第5期10-13,共4页
Applied Science and Technology
基金
黑龙江省自然科学基金资助项目(F201339)
关键词
模式分类
近邻原则
K-近邻
k-近质心近邻
局部权重
pattern classification
nearest neighbor rule
k-nearest neighbor rule
k-nearest centroid neighbor rule
local weight
