《人工智能利用BP网络对26个字母进行识别.docx》由会员分享，可在线阅读，更多相关《人工智能利用BP网络对26个字母进行识别.docx（5页珍藏版）》请在金锄头文库上搜索。
1、 人工智能利用BP网络对26个字母进行识别（matlab） clear;%清除工作区中的变量alphabet,targets = prprob;%通过prprob函数得到训练字母集和目标集R,Q = size(alphabet);%取得字母集的维数S2,Q = size(targets);%取得目标集的维数，在这里目标集是26个二进制数。分别代表AZS1 = 10;%隐含层个数net = newff(minmax(alphabet),S1 S2,logsig logsig,traingdx);%创建神经网络，两层，函数为Logsig。net.LW2,1 = net.LW2,1*0.01;%设定
2、权值net.b2 = net.b2*0.01;%设定偏差向量net.performFcn = sse; %运用和平方最小误差方式net.trainParam.goal = 0.1;%训练精度目标为0.1 net.trainParam.epochs = 5000; %迭代5000次net.trainParam.mc = 0.95; %冲量为0.95P = alphabet;%训练集T = targets;%目标集net,tr = train(net,P,T);%训练网络netn = net;%创建带有噪声的网络netn.trainParam.goal = 0.6; %训练精度目标为0.6 net
3、n.trainParam.epochs = 300; %迭代300次T = targets targets targets targets;%扩展目标集for pass = 1:10 fprintf(Pass = %.0fn,pass); P = alphabet, alphabet, . (alphabet + randn(R,Q)*0.1), . (alphabet + randn(R,Q)*0.2);%将训练集加入噪声 netn,tr = train(netn,P,T);%训练带有噪声的神经网络endnetn.trainParam.goal = 0.1; %训练精度目标为0.1 netn
4、.trainParam.epochs = 500; %迭代500次P = alphabet;%重设训练集T = targets;%重设目标集netn,tr = train(netn,P,T);%训练代有噪声的神经网络noise_range = 0:.05:.5;%设定噪声max_test = 100;network1 = ;network2 = ;T = targets;%设定目标集for noiselevel = noise_range fprintf(Testing networks with noise level of %.2f.n,noiselevel); errors1 = 0;
5、errors2 = 0; for i=1:max_test P = alphabet + randn(35,26)*noiselevel;%仿真集加入噪声 A = sim(net,P);%对用无噪声训练的网络进行仿真 AA = compet(A);%输出为1的那个对应输入向量为竞争获胜的输入向量 errors1 = errors1 + sum(sum(abs(AA-T)/2;%计算误差 An = sim(netn,P);%对用有噪声训练的网络进行仿真 AAn = compet(An);%输出为1的那个对应输入向量为竞争获胜的输入向量 errors2 = errors2 + sum(sum(ab
6、s(AAn-T)/ 2;%计算误差 end network1 = network1 errors1/26/100;%计算出Network1的误差 network2 = network2 errors2/26/100;%计算出Network2的误差endplot(noise_range,network1*100,-,noise_range,network2*100,*);%打出效果图title(误差百分数);xlabel(噪声水平);ylabel(Network 1 - - Network 2 *);noisyF=alphabet(:,6)+randn(35,1)*0.2;plotchar(no
7、isyF)A1= sim(net,noisyF);plotchar(noisyF) ; plotchar(noisyF) ;figure;A = sim(net,P); A1= sim(net,noisyF); A1=compet(A2); A1=compet(A1); plotchar(noisyF);figure;A2= sim(netn,noisyF); A2=compet(A2); plotchar(noisyF); figure plotchar(A2)figure plotchar(A1);TRAINGDX, Epoch 0/5000, SSE 168.447/0.1, Gradi
8、ent 46.0382/1e-006TRAINGDX, Epoch 25/5000, SSE 24.9713/0.1, Gradient 0.563004/1e-006TRAINGDX, Epoch 50/5000, SSE 25.4485/0.1, Gradient 0.363459/1e-006TRAINGDX, Epoch 75/5000, SSE 25.532/0.1, Gradient 0.333604/1e-006TRAINGDX, Epoch 100/5000, SSE 25.4204/0.1, Gradient 0.407/1e-006TRAINGDX, Epoch 125/5
9、000, SSE 24.5179/0.1, Gradient 0.590354/1e-006TRAINGDX, Epoch 150/5000, SSE 20.1153/0.1, Gradient 1.03697/1e-006TRAINGDX, Epoch 175/5000, SSE 3.3111/0.1, Gradient 0.741009/1e-006TRAINGDX, Epoch 199/5000, SSE 0.0903872/0.1, Gradient 0.0562929/1e-006TRAINGDX, Performance goal met.Pass = 1TRAINGDX, Epo
10、ch 0/300, SSE 4.38727/0.6, Gradient 3.91707/1e-006TRAINGDX, Epoch 25/300, SSE 2.34149/0.6, Gradient 1.61314/1e-006TRAINGDX, Epoch 50/300, SSE 1.66179/0.6, Gradient 0.87181/1e-006TRAINGDX, Epoch 75/300, SSE 0.860046/0.6, Gradient 0.390595/1e-006TRAINGDX, Epoch 89/300, SSE 0.595819/0.6, Gradient 0.201
11、143/1e-006TRAINGDX, Performance goal met.Pass = 2TRAINGDX, Epoch 0/300, SSE 1.69802/0.6, Gradient 2.60629/1e-006TRAINGDX, Epoch 25/300, SSE 1.0446/0.6, Gradient 0.670983/1e-006TRAINGDX, Epoch 50/300, SSE 0.809773/0.6, Gradient 0.387906/1e-006TRAINGDX, Epoch 73/300, SSE 0.597462/0.6, Gradient 0.19625
12、7/1e-006TRAINGDX, Performance goal met.Pass = 3TRAINGDX, Epoch 0/300, SSE 1.84056/0.6, Gradient 3.45909/1e-006TRAINGDX, Epoch 25/300, SSE 1.03318/0.6, Gradient 0.911047/1e-006TRAINGDX, Epoch 50/300, SSE 0.706745/0.6, Gradient 0.415866/1e-006TRAINGDX, Epoch 60/300, SSE 0.599226/0.6, Gradient 0.295716
13、/1e-006TRAINGDX, Performance goal met.Pass = 4TRAINGDX, Epoch 0/300, SSE 1.60648/0.6, Gradient 2.3488/1e-006TRAINGDX, Epoch 25/300, SSE 1.01923/0.6, Gradient 0.815083/1e-006TRAINGDX, Epoch 50/300, SSE 0.784938/0.6, Gradient 0.396484/1e-006TRAINGDX, Epoch 74/300, SSE 0.592602/0.6, Gradient 0.188876/1
14、e-006TRAINGDX, Performance goal met.Pass = 5TRAINGDX, Epoch 0/300, SSE 3.77602/0.6, Gradient 4.17101/1e-006TRAINGDX, Epoch 25/300, SSE 1.53422/0.6, Gradient 2.16367/1e-006TRAINGDX, Epoch 50/300, SSE 0.825853/0.6, Gradient 0.63716/1e-006TRAINGDX, Epoch 71/300, SSE 0.594703/0.6, Gradient 0.226427/1e-0
15、06TRAINGDX, Performance goal met.Pass = 6TRAINGDX, Epoch 0/300, SSE 1.42439/0.6, Gradient 2.20268/1e-006TRAINGDX, Epoch 25/300, SSE 0.834845/0.6, Gradient 0.700766/1e-006TRAINGDX, Epoch 50/300, SSE 0.625815/0.6, Gradient 0.35906/1e-006TRAINGDX, Epoch 54/300, SSE 0.594342/0.6, Gradient 0.318662/1e-006TRAINGDX, Performance goal met.Pass = 7TRAINGDX, Epoch 0/300, SSE 2.09651/0.6, Gradie
