核函数是利用支持向量机解决不可分问题时引入的一种非线性变换的手段。基本思想是通过非线性变换，使样本变换之后的特征空间中变得线性可分。然后利用线性可分时构造最优超平面的方法，在特征空间中实现最优超平面的求解。-Kernel function is the use of support vector machine to resolve the issue can not be separated from the introduction of a nonlinear transform mean

张学工老师模式识别第四次作业，分别用C均值方法，分层聚类方法和非监督参数下正态分布函数估计的贝叶斯决策对身高体重二维数据进行分类。-Zhang engineering teacher fourth pattern recognition operations, respectively C-means, hierarchical clustering methods, and non-normal distribution function under the supervision of th

Original Non-local means filtering of an image, searching extent is the whole image