Question

matlab bp神经网络预测问题

load('P1.mat');%1920*6矩阵
load('T.mat');%1920*1矩阵
load('Y.mat');%480*6矩阵
load('E1.mat');%480*1矩阵
net=newff(minmax(P),[10, 5], {'tansig' 'purelin' 'purelin'}, 'traingdm');
net.trainParam.epochs=2500;
net.trainParam.goal=0.01;
net.trainParam.show=10;
net.trainParam.lr=0.5;
net.trainParam.mc=0.6;
net.train(net,P,T);
Y= sim(net, Y); %仿真
plot(E,'o');
hold on
plot(Y,'-');%结果对比

结果是：
Warning: NEWFF used in an obsolete way.
> In obs_use at 18
In newff>create_network at 127
In newff at 102
In fenduan at 9
See help for NEWFF to update calls to the new argument list.

Maximum variable size allowed by the program is exceeded.

Error in nnMex2.codeHints (line 117)
hints.TEMP = zeros(1,ceil(tempSize/8),'double');

Error in nncalc.setup2 (line 13)
calcHints = calcMode.codeHints(calcHints);

Error in network/sim (line 321)
[calcLib,calcNet] = nncalc.setup2(calcMode,calcNet,calcData,calcHints,isParallel);

Error in fenduan (line 16)
Y= sim(net, Y); %仿真

程序该怎么改？求高手。

Answer 1

-请问此处如何处理，能够达到让pn等于所有的已知样本来作为预测样本输入????

不知道问题出在哪里，为什么不能得到我想要的下一组数据呢？矛盾啊

http://wenku.baidu.com/link?url=5sPeHwbbotF80NS4oh94zeMshhP68JhbRIqTcBeeAfJ9ZnVheg-_W07sq789GvER80S4IieJ8VRjoh7y-txeT1nFpw5Wi4YX9c9PaogxfOS

追问

那我现在程序出现什么问题？

Answer 2

我接触matlab12年了，第一次见不用训练直接sim的bp网络，他报的信息是你没有按照常规用神经网络！

追问

已经找到解答了，谢谢啊。