基于麻雀算法优化BP神经网络的多分类识别，基于麻雀算法改进BP神经网络的多分类预测代码，ssa-bp多分类代码

clc clear close all warning off rng('default') load maydata.mat m1=5; m =720; rng(12) n=randperm(800); %节点个数 inputnum=5; hiddennum=10; outputnum=1; m = round(0.9*length(num)); n = randperm(length(num)); %训练数据和预测数据 input_train=num(n(1:m),1:m1)'; input_test=num(n(m+1:end),1:m1)'; output_train=num(n(1:m),m1+1)'; output_test=num(n(m+1:end),m1+1)'; %选连样本输入输出数据归一化 [inputn,inputps]=mapminmax(input_train); [outputn,outputps]=mapminmax(output_train); %构建网络 % net=newff(inputn,outputn,hiddennum); net = newff(minmax(inputn),[10,1],{'logsig','tansig'},'trainlm'); dim=inputnum*hiddennum+hiddennum+hiddennum*outputnum+outputnum; maxgen=30; % 进化次数 sizepop=20; %种群规模 popmax=5; popmin=-5; %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% P_percent = 0.2; % The population size of producers accounts for "P_percent" percent of the total population size pNum = round( sizepop * P_percent ); % The population size of the producers %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% for i=1:sizepop pop(i,:)=5*rands(1,dim); % V(i,:)=rands(1,21); fitness(i)=fun(pop(i,:),inputnum,hiddennum,outputnum,net,inputn,outputn); end %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% pFit = fitness; [ fMin, bestI ] = min( fitness ); % fMin denotes the global optimum fitness value bestX = pop( bestI, : ); % bestX denotes the global optimum position corresponding to fMin %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%麻雀搜索算法 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% for t = 1 : maxgen [ ans, sortIndex ] = sort( pFit );% Sort. [fmax,B]=max( pFit ); worse= pop(B,:); r2=rand(1); if(r2<0.8) for i = 1 : pNum % Equation (3) r1=rand(1); pop( sortIndex( i ), : ) = pop( sortIndex( i ), : )*exp(-(i)/(r1*maxgen)); fitness(sortIndex( i ))=fun(pop(sortIndex( i ),:),inputnum,hiddennum,outputnum,net,inputn,outputn); end else for i = 1 : pNum pop( sortIndex( i ), : ) = pop( sortIndex( i ), : )+randn(1)*ones(1,dim); fitness(sortIndex( i ))=fun(pop(sortIndex( i ),:),inputnum,hiddennum,outputnum,net,inputn,outputn); end end [ fMMin, bestII ] = min( fitness ); bestXX = pop( bestII, : ); for i = ( pNum + 1 ) : sizepop % Equation (4) A=floor(rand(1,dim)*2)*2-1; if( i>(sizepop/2)) pop( sortIndex(i ), : )=randn(1)*exp((worse-pop( sortIndex( i ), : ))/(i)^2); else pop( sortIndex( i ), : )=bestXX+(abs(( pop( sortIndex( i ), : )-bestXX)))*(A'*(A*A')^(-1))*ones(1,dim); end fitness(sortIndex( i ))=fun(pop(sortIndex( i ),:),inputnum,hiddennum,outputnum,net,inputn,outputn); end c=randperm(numel(sortIndex)); b=sortIndex(c(1:3)); for j = 1 : length(b) % Equation (5) if( pFit( sortIndex( b(j) ) )>(fMin) ) pop( sortIndex( b(j) ), : )=bestX+(randn(1,dim)).*(abs(( pop( sortIndex( b(j) ), : ) -bestX))); else pop( sortIndex( b(j) ), : ) =pop( sortIndex( b(j) ), : )+(2*rand(1)-1)*(abs(pop( sortIndex( b(j) ), : )-worse))/ ( pFit( sortIndex( b(j) ) )-fmax+1e-50); end fitness(sortIndex( i ))=fun(pop(sortIndex( i ),:),inputnum,hiddennum,outputnum,net,inputn,outputn); end %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% for i = 1 : sizepop if ( fitness( i ) < pFit( i ) ) pFit( i ) = fitness( i ); pop(i,:) = pop(i,:); end if( pFit( i ) < fMin ) fMin= pFit( i ); bestX =pop( i, : ); end end yy(t)=fMin; end %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% %% 迭代寻优 x=bestX %% 结果分析 plot(yy) title(['适应度曲线 ' '终止代数＝' num2str(maxgen)]); xlabel('进化代数');ylabel('适应度'); ylim([min(yy)*0.8 max(yy)*1.2]) %% 把最优初始阀值权值赋予网络预测 % %用麻雀搜索算法优化的BP网络进行值预测 w1=x(1:inputnum*hiddennum); B1=x(inputnum*hiddennum+1:inputnum*hiddennum+hiddennum); w2=x(inputnum*hiddennum+hiddennum+1:inputnum*hiddennum+hiddennum+hiddennum*outputnum); B2=x(inputnum*hiddennum+hiddennum+hiddennum*outputnum+1:inputnum*hiddennum+hiddennum+hiddennum*outputnum+outputnum); net.iw{1,1}=reshape(w1,hiddennum,inputnum); net.lw{2,1}=reshape(w2,outputnum,hiddennum); net.b{1}=reshape(B1,hiddennum,1); net.b{2}=B2; %% 训练 %网络进化参数 net.trainParam.epochs=5000; net.trainParam.lr=0.1; net.trainParam.goal=0.00001; %网络训练 [net,tr]=train(net,inputn,outputn); %%预测 %数据归一化 inputn_test=mapminmax('apply',input_test,inputps); input_test1=[20000 471 63 1205 1677]'; inputn_test1=mapminmax('apply',input_test1,inputps); an=sim(net,inputn_test); an1=sim(net,inputn_test1); test_simu=round(mapminmax('reverse',an,outputps)); test_simu1=round(mapminmax('reverse',an1,outputps)); error=test_simu-output_test; figure plot(test_simu,'r-o') hold on plot(output_test,'k-*') hold off xlabel('样本') ylim([0.5 5.5]) set(gca,'fontsize',12) legend('ssa-bp','实际值') [output_test1,my] = sort(output_test); test_simu1 = test_simu(my); figure plot(test_simu1(1,:),'r-o')%预测的结果数据画图：代表虚线，O代表圆圈标识，G代表绿色 hold on plot(output_test1(1,:),'k-*');%期望数据，即真实的数据画图，-代表实现，*就是代表*的标识 hold on legend('SSA-BP预测输出','期望输出')%标签 title('BP神经网络','fontsize',12)%标题 字体大小为12 xlabel('样本') ylim([0.5 5.5]) set(gca,'fontsize',12) figure plot(error) title('仿真预测误差','fontsize',12); xlabel('','fontsize',12); ylabel('误差','fontsize',12); acc = sum(output_test==test_simu)/length(output_test) [cfmat,order] = confusionmat(output_test,test_simu); mat=cfmat; k=5; %% 混淆矩阵制图 [cfmat,order] = confusionmat(output_test,test_simu); mat=cfmat; k=5; figure xx = [ones(256,1) ones(256,1) [255:-100/255: 155 ]'/255]; imagesc(mat); %# Create a colored plot of the matrix values colormap(xx); %flipud(summer) # Change the colormap to gray (so higher values are title('SSA-BP混淆矩阵'); textStrings = num2str(mat(:),'%0.02f'); %# Create strings from the matrix values textStrings = strtrim(cellstr(textStrings)); %# Remove any space padding [x,y] = meshgrid(1:k); hStrings=text(x(:),y(:),textStrings(:),'HorizontalAlignment','center'); midValue = mean(get(gca,'CLim')); %# Get the middle value of the color range textColors = repmat(mat(:) > midValue,1,3); %set(hStrings,{'Color'},num2cell(textColors,2)); %# Change the text colors锛? set(gca,'XTick',1:5,... 'XTickLabel',{'1','2','3','4','5'},... %# and tick labels 'YTick',1:5,... %鍚屼笂 'YTickLabel',{'1','2','3','4','5'},... 'TickLength',[0 0]);