load spectra_data.mat
tic
temp = randperm(size(NIR,1));
P_train = NIR(temp(1:50),:)';
T_train = octane(temp(1:50),:)';
P_test = NIR(temp(51:end),:)';
T_test = octane(temp(51:end),:)';
N = size(P_test,2);
[p_train, ps_input] = mapminmax(P_train,0,1);
p_test = mapminmax('apply',P_test,ps_input);
[t_train, ps_output] = mapminmax(T_train,0,1);
R1=0;
I=0;
J=0;
for i=9:10
for j=5:6
net = newff(p_train,t_train,[i,j]);
net.trainParam.epochs = 1000; %迭代次数
net.trainParam.goal = 1e-3; %mse均方根误差小于这个值训练结束
net.trainParam.lr = 0.01; %学习率
net = train(net,p_train,t_train);
t_sim = sim(net,p_test);
T_sim = mapminmax('reverse',t_sim,ps_output);
error = abs(T_sim - T_test)./T_test;
R2 = (N * sum(T_sim .* T_test) - sum(T_sim) * sum(T_test))^2 / ((N * sum((T_sim).^2) - (sum(T_sim))^2) * (N * sum((T_test).^2) - (sum(T_test))^2));
if (j==9&&i==5)
R1=R2;
I=i;
J=j;
continue;
else
if (R2>R1)
R1=R2;
t_sim_final=t_sim;
T_sim_final=T_sim;
I=i;
J=j;
else
continue
end
end
end
end
result = [T_test' T_sim_final' error']
toc
figure
plot(1:N,T_test,'b:*',1:N,T_sim_final,'r-o')
legend('真实值','预测值')
xlabel('预测样本')
ylabel('辛烷值')
string = {'测试集辛烷值含量预测结果对比';['R^2=' num2str(R1)]};
title(string)
figure
plot(1:N,error,'g:*')
