求Matlab指导，Undefined function 'minimize' for input arguments of type 'struct'.错误是

clear all
x(1)=1.02;
x(2)=1.56;
col = {'k',[.8,0,0],[0,.5,0],'b',[0,.75,.75],[.7,0,.5]};
sn = 0.2;
for i=3:100
x(i)=0.75*x(i-1)+0.25*x(i-2)+randn(1,1)*sn;
end
%i=1:100;
%plot(i,x,'r*');
L=2;
xtr=ones(98,2);
for i=1:98
for ii=1:L
xtr(i,ii)=x(i);
end
end
for t=1:98
yt(t)=x(t+2);
end
ytr=yt';
cov = {@covNNone}; ell = 0.4; sf = 1; % setup the GP
hyp0.cov = log([ell;sf]);
mean = {@meanSum,{@meanLinear,@meanConst}};a = [1/5;1/5]; b =1; % m(x) = a*x+b
hyp0.mean = [a;b];
lik={@likGauss};
inf={@infExact};
hyp0.lik = log(sn);
Ncg = 50; % number of conjugate gradient steps
sdscale = 0.5; % how many sd wide should the error bars become?
hyp = minimize(hyp0,'gp', -Ncg, inf, mean, cov, lik, xtr, ytr); % opt hypers
[nlZ] = gp(hyp, inf, mean, cov, lik, xtr, ytr);
[ymu, ys2] = gp(hyp, inf, mean, cov, lik, xtr, ytr, xtr); % predict
i=1:100;
figure;
hold on;
plot(i,x,'r*');
i=3:100;
plot(i,ymu,'B+');
ysd = sdscale*sqrt(ys2);
plot(i,ymu+ysd,'k');
plot(i,ymu-ysd,'k');
%fill([i,flipud(i)],[ymu+ysd;flipud(ymu-ysd)],'b');
k=1;
xi=i;
ytr1=(ymu+ysd)';
ytr2=(ymu-ysd)';
for t=0:length(xi)-1
xi1(k)=xi(length(xi)-t);
Z2(k)=ytr2(length(xi)-t);
k=k+1;
end
fill([xi xi1],[ytr1 Z2],col{2},'EdgeColor',col{2},'FaceAlpha',0.1,'EdgeAlpha',0.3) % fill

你可以把Carl Edward Rasmussen高斯过程MATLAB工具包里的minimize函数，拷贝到你的文件夹里。

minimize函数不是matlab中已有的函数，应先定义为自定义函数

楼主高斯过程处理的怎么样了 能不能交流一下 我遇到的问题和你很相似

争议你可以把Carl Edward Rasmussen高斯过程MATLAB工具包里的minimize函数，拷贝到你的文件夹里。 再看看别人怎么说的。