principal component analysis Files

This program demonstrates the application of principal component analysis on hand-writing recognition. The related data is stored in mnistAll.mat.
The program uses minimum-distance classifier to discriminate. The better way is to use more advanced methods such as Fisher's multi-class criterion
The correct rate is around 0.7.

The program is created by Pengfei Chen on Nov. 13, 2013.
Copyright (C) 2013 by Pengfei Chen
jiyuanyizhong@hotmail.com