人工神经网络matlab程序.docVIP

神经网络MATLAB程序 %根据预测方法得到输入向量和目标向量 P=[0.4413 0.4707 0.6953 0.8133 0.4379 0.4677 0.6981 0.8002 0.4517 0.4725 0.7006 0.8201; 0.4379 0.4677 0.6981 0.8002 0.4517 0.4725 0.7006 0.8201 0.4557 0.4790 0.7019 0.8211; 0.4517 0.4725 0.7006 0.8201 0.4557 0.4790 0.7019 0.8211 0.4601 0.4811 0.7101 0.8298;]; T=[0.4557 0.4790 0.7019 0.8211; 0.4601 0.4811 0.7101 0.8298; 0.4612 0.4845 0.7188 0.8312]; %输入向量的取值范围为[0 1]，用threshold来标记 threshold=[0 1;0 1;0 1;0 1;0 1;0 1;0 1;0 1;0 1;0 1;0 1;0 1]; %创建一个Elman神经网络，隐含层的神经元个数为17个，4个输出层神经元，隐含层激活函 %数为tansig，输出层激活函数为purelin net=newelm(threshold,[17,4],{tansig,purelin}); net.trainParam.epochs=3000; net=init(net); net=train(net,P,T); %输入测试数据 P_test=[0.4557 0.4790 0.7019 0.8211 0.4601 0.4811 0.7101 0.8298 0.4612 0.4845 0.7188 0.8312]; T_test=[0.4615 0.4891 0.7201 0.8330]; y=sim(net,P_test) %在测试数据下，计算网络的输出和实际目标向量之间的差值 error=y-T_test; %在坐标平面上画出差值曲线 plot(1:4,error,-); 程序运行后，结果如图2-29所示，命令行窗口中的结果如下： …… TRAINGDX, Epoch 2950/3000, MSE 2.65822e-005/0, Gradient 0.000916222/1e-006 TRAINGDX, Epoch 2975/3000, MSE 2.64788e-005/0, Gradient 0.000221814/1e-006 TRAINGDX, Epoch 3000/3000, MSE 2.6293e-005/0, Gradient 0.000102435/1e-006 TRAINGDX, Maximum epoch reached, performance goal was not met. y = 0.4761 0.4916 0.7217 0.8344 Hopfield %数字1的点阵表示 one=[-1 -1 -1 1 1 1 1 -1 -1 -1 -1 -1 -1 1 1 1 1 -1 -1 -1 -1 -1 -1 1 1 1 1 -1 -1 -1 -1 -1 -1 1 1 1 1 -1 -1 -1 -1 -1 -1 1 1 1 1 -1 -1 -1 -1 -1 -1 1 1 1 1 -1 -1 -1 -1 -1 -1 1 1 1 1 -1 -1 -1 -1 -1 -1 1 1 1 1 -1 -1 -1 -1 -1 -1 1 1 1 1 -1 -1 -1 -1 -1 -1 1 1 1 1 -1 -1 -1]; %数字2的点阵表示 two=[1 1 1 1 1 1 1 1 -1 -1 1 1 1 1 1 1 1 1 -1 -1 -1 -1 -1 -1 -1 -1 1 1 -1 -1 -1 -1 -1 -1 -1 -1 1 1 -1 -1 1 1 1 1 1 1 1 1 -1 -1 1 1 1 1 1 1 1 1 -1 -1 1 1 -1 -1 -1 -1 -1 -1 -1 -1 1 1 -1 -1 -1 -1 -1 -1 -1 -1 1 1 1 1 1 1 1 1 -1 -1 1 1 1 1 1 1 1 1 -1 -1]; %设定网络的目标向量 T=[one;two]; %创建一个Hopfield神经网络 net=newhop(T); %给定一个受噪声污染的数字2的点阵，所谓噪声是指数字点阵中某些本应为1的方块变成了-1 no2={[1 1 1 -1 1 1 -1 1 -1 -1 1 1 1