此为模式识别中的Fisher线性判别方法求分类器。进行了男、女错误率和总错误率的统计。

全部程序流程如下：

1、读取FAMALE.TXT文件把身高或体重给数组，并求x1的样本均值m1和内离散度S1；

2、读取MALE.TXT文件把身高或体重给数组，并求x2的样本均值m2和内离散度S2；

3、求最佳变换向量和阀值点；

4、读取Test2.txt文件把对应的身高或体重给数组A并求A的样本数M；

5、把最佳变换向量和阀值点、x代入判别方程g（x）；

6、对本判别的错误率进行统计。-This is the pattern recognition method, to calculate the Fisher linear discriminant classifier .Were male , female and total error rate error rate statistics .

All program flow is as follows :

1 , read FAMALE.TXT file to height or weight to the array, and find the sample mean m1 x1 and internal dispersion S1

2 , read MALE.TXT file to height or weight to the array, and find the sample mean m2 and x2 within the dispersion S2

3 , for the best transformation vector and the threshold point　

4 , read Test2.txt file to the corresponding height or weight requirements for the array A and A is the number of samples M

5 , the best transformation vector and the threshold point , x substituted into the discriminant equation g (x)

6 , determine the error rate of the statistics .