www.gusucode.com > 通过用最小距离分类判别方法，用MATLAB程序找出最小距离分类判别时的识别界面，从而进行识别已知的两类训练样本，并分析其识别错误率 > 通过用最小距离分类判别方法，用MATLAB程序找出最小距离分类判别时的识别界面，从而进行识别已知的两类训练样本，并分析其识别错误率/新建文件夹/zuoye.m

    w1=[[2,0];[2,2];[2,4];[3,3]];
w1(:,1);
w2=[[0,3];[-2,2];[-1,-1];[1,-2];[3,-1]];
n1=length(w1);
n2=length(w2);
w1_t=0;
for i=1:n1
 w1_t=w1_t+w1(i,:);
end
w1_e=w1_t./n1;
w2_t=0;
for i=1:n1
 w2_t=w2_t+w2(i,:);
end
w2_e=w2_t./n1;
w12mid=(w1_e+w2_e)./2;
w12mid(1);
%slop=-(w2_e(2)-w1_e(2))./(w2_e(1)-w1_e(1));
%t=10;
%a=[w12mid(1)-t,w12mid(1),w12mid(1)+t]';
%b=[w12mid(2)-t.*slop,w12mid(1),w12mid(1)+t.*slop]';
%figure(3)
%line(a,b);
%hold on
plot(w1(:,1)',w1(:,2)','ro');
hold on
plot(w1_e(:,1),w1_e(:,2),'k*');
hold on
plot(w2(:,1)',w2(:,2)','gx');
hold on
plot(w2_e(:,1),w2_e(:,2),'k*');
hold on
slop=-(w2_e(2)-w1_e(2))./(w2_e(1)-w1_e(1));
t=10;
a=[w12mid(1)-t,w12mid(1),w12mid(1)+t]';
b=[w12mid(2)-t.*slop,w12mid(1),w12mid(1)+t.*slop]';
line(a,b);
hold on
hold off