用其中一半的数据采用ANN-BP算法设计分类器，另一半数据用于测试分类器性能。-Half of them with the data ANN-BP algorithm design classifier, and the other half data used to test the performance of classification.

线性分类器设计，1_k近邻法，剪辑法，ANN-BP神经网络法，聚类分析法，特征选择，特征提取等模式识别常用算法，内有matlab可运行实现，还有word文档的说明。对于学习，是做好的资料。-Commonly used algorithm for linear classifier design the 1_k nearest neighbor, clip, ANN-BP neural network method, cluster analysis, feature selection, fea

一个运用BP神经网络设计数字识别的分类器的demo，用opencv实现的。-A digit recognition using BP neural network design classifier demo, with opencv.