摘要
传统的降维方法有主元分析法、多维尺度法,但是此方法对于线性结构降维效果较好,而对于非线性的高维特征处理效果差强人意。等距离映射法通过原始数据与降维数据之间的关系实现降维处理,本文运用KNN加权来实现分类决策。
The traditional method of dimension reduction with principal component analysis, multidimensional scaling method. But this method for linear dimensionality reduction effect is good, and for the high dimension nonlinear treatment effect is just passable. Based on the relationship between the original data and the dimension reduction data, the method of equal distance mapping was used to reduce the dimension of the data. This paper uses KNN weighting to realize the classification decision.
引文
[1]闫德勤,刘胜蓝.基于局部切空间偏离度的自适应邻域选取算法[J].模式识别与人工智能,2010,23(6):815-821.
[2]张妮,田学民.基于等距离映射的非线性动态故障检测方法[J].上海交通大学学报,2011(8):1202-1206.
[3]朱明旱,罗大庸,易励群,等.一种序列的加权KNN分类方法[J].电子学报,2009,37(11):2584-2588.