matlab mda,Matlab的学习最大似然估计，PCA和MDA的一段代码

一道Pattern Recognition作业题，觉得对学习最大似然估计，PCA和MDA有所帮助，就贴上来。

程序是自己写的，新手，可能编的不太好，还请各位看官指教，谢谢。

IMOX dataset has been used in pattern recognition literature to evaluate the performance of various feature selection, classification and clustering algorithms. This dataset consists of 192 8-dimensional patterns belonging to four character classes (I, M, O and X). There are 48 patterns per class. The 8 features correspond to the distance of a character to the:



upper left boundary



lower right boundary



upper right boundary



lower left boundary



middle left boundary



middle right boundary



middle upper boundary



middle lower boundary

The class labels are indicated at the end of every pattern.

(a) Assume that each pattern class has a multivariate Gaussian density with unknown parameters. Design a maximum-likelihood classifier for this problem. Randomly choose half the data (from each class) for training the classifier and the remaining half for testing the classifier. Repeat this data-splitting ten times.

Report the average error rate of the quadratic classifier.

(b) Use Principal Component Analysis to get 2- and 3-dimensional representations (projection) of the data. Label the patterns according to the category information in the projection. How well are the classes separated in these projections?

(c) Use Multiple Discriminant Analysis to get 2- and 3-dimensional projections of the data. Label the patterns according to the category information in these projections. Are the pattern classes better separated in these plots than in (b).

code.rar

(4.63 KB, 下载次数: 26791)

2009-11-2 19:47 上传

点击文件名下载附件