# matlab代做highspeedlogic★动态模糊神经网络

### 时间：2015-10-5 18:28:22 点击：

核心提示：matlab代做，FPGA代做，淘宝，专业代做MATLAB、FPGA博士/硕士/本科毕业设计、项目仿真、Coursework、Assignment ...

QQ: 1224848052

function y=mdist(p,c,b)
% This program is used to compute Mahalanobis distance
% p--r*q matrix,which is the incoming sample data
% c--r*s matrix, which is the center matrix
% b--r*s matrix, which is the width matrix
% Revised 11-5-2006
[r,q]=size(p);
[r,s]=size(c);
[r,s]=size(b);
y=zeros(s,q);
if r==1
for i=1:s
x=c(:,i)*ones(1,q);
d=abs(p-x);
xx=b(:,i)*ones(1,q);
dd=d./xx;
y(i,:)=dd.*dd;
end
else
for i=1:s
x=c(:,i)*ones(1,q);
d=abs(p-x);
xx=b(:,i)*ones(1,q);
dd=d./xx;
y(i,:)=sum(dd.*dd);
end
end

function [w,y,a]=mweight(in,c,wb,out)
% This program is used to adjust the weights of TSK FNNs based
% on simplified Mahalanobis
% Revised 11-5-2006
[r,n]=size(in);
[r,u]=size(c);
[r,u]=size(wb);
[s2,n]=size(out);
a=exp(-mdist(in,c,wb));
a1=transf(a,in);
w=out/a1;
y=w*a1;

function [w,a]=orthogonalize(p)
[u,v]=size(p);
w(:,1)=p(:,1);
a=eye(v);
for k=2:v
b=zeros(u,1);
for i=1:k-1
a(i,k)=w(:,i)'*p(:,k)/(w(:,i)'*w(:,i));
b=b+a(i,k)*w(:,i);
end

w(:,k)=p(:,k)-b;
end

• 百度搜索
• 查阅资料过程中
• 论坛发现
• 百度贴吧发现
• 朋友介绍
• 没有相关文章
• 大名：
• 内容：