DEMO

clc;
clear;
close all;
load ..\CFCC_mat\all_data_label.mat
% load all_data_label_02.mat
%matlab神经网络工具箱权值和阈值每次都是随机初始化的，加上"setdemorandstream(pi)"即可固定随机种子，使其训练结果不变
setdemorandstream(pi);train_y=train_y';
test_y=test_y'; 
P=size(test_x);
w=size(train_x);
%%category=26;%26种类别
M=26*10;%测试用样本数
bp_number=80; %隐层神经元个数
% bp_number =round( log( size(train_x,2) )/ log(2)  );
input_label_test=zeros(1,M); 
% output_label=zeros(1,M); 
%%
for j=1:P(1) %P（1）=90input_label_test(j)= find(test_y(:,j))    ;      %正确，返回T1中第j列不为0元素的行数end%%
%特征值归一化; 
[input,minI,maxI] = premnmx( train_x')  ;%input归一化矩阵，minI，每行最小值，maxI，每行最大值
%for i=10:300
%创建神经网络，隐藏层与输出层对应的传输转移函数分别为"tansig","logsig"%[24,9]
net=newff(minmax(input),[bp_number,category],{'tansig' 'logsig'} ,'traingdx');
% 'traingdx' 自适应调整学习速率附加动量因子梯度下降反向传播算法训练函数
net.trainParam.lr=0.001;%学习率
net.trainParam.show=5000;%显示中间结果的周期
net.trainParam.epochs=20000;%最大迭代次数
net.trainParam.goal=0.0001; %神经网络训练的目标误差%%开始训练
net=train(net,input,train_y);  %T9x900 
save trainNet net minI %%%train
trainInput = tramnmx ( train_x' , minI, maxI ) ;
TT=sim(net,trainInput);%返回测试样本的输出预测值
error2 = train_y - round(TT);
RX=0;
for s=1:w(1)if(sum(error2(:,s))==0)RX = RX + 1;   endendTrainidentifyrate = RX/(w(1)) %%
% 测试数据归一化 test
testInput = tramnmx ( test_x' , minI, maxI ) ;
RS=sim(net,testInput);%test_input 400xM,每一列为一个样本 RS9x90，9类一共M个样本% U=round(RS);
[~,output_label_test] = max(RS,[],1); %返回每列最大值所处行数%%
%视图化
%  test_labels  =input_label_test';
%  predict_label=output_label_test';figure;
hold on;plot(input_label_test','o');   %test_labels 
plot(output_label_test','r+');%predict_label xlabel('测试集样本','FontSize',12);
ylabel('类别标签','FontSize',12);
legend('实际测试集分类','预测测试集分类');
title('测试集的实际分类和预测分类图','FontSize',12);
%设置x轴范围和刻度
set(gca,'XLim',[0 260]);%X轴的数据显示范围
set(gca,'XTick',[0:10:260]);%设置要显示坐标刻度
% set(gca,'XTickLabel',[0:1:10]);%给坐标加标签 %设置y轴范围和刻度set(gca,'YLim',[1 26]);%X轴的数据显示范围
set(gca,'YTick',[0:1:26]);%设置要显示坐标刻度
% set(gca,'YTickLabel',[95:1:101]);%给坐标加标签 
grid on;%% III 计算准确率c=input_label_test-output_label_test;number=0;error_label_number =1;for i=1:1:Mif(c(1,i)==0)number = number+1;end
%      if (abs(c(1,i))==1 ) %不等于
%          error_label(error_label_number,1)= i; %误差为1错误标签位置
%          error_label_number = error_label_number+1;
%      end
%if (abs(c(1,i))~=0 ) %不等于error_location(error_label_number,1)= i; %误差为1错误标签位置error_label_number = error_label_number+1;endendaccuracy = number/M
save error_label.mat  error_location%%
data =[input_label_test',output_label_test'];
data =transform(data);
figure;
hold on;
plot(data(:,1)','o');   %test_labels 
plot(data(:,2)','r+');%predict_label xlabel('测试集样本','FontSize',12);
ylabel('类别标签','FontSize',12);
legend('实际测试集分类','预测测试集分类');
title('测试集的实际分类和预测分类图','FontSize',12);
%设置x轴范围和刻度
set(gca,'XLim',[0 260]);%X轴的数据显示范围
set(gca,'XTick',[0:10:260]);%设置要显示坐标刻度
% set(gca,'XTickLabel',[0:1:10]);%给坐标加标签 %设置y轴范围和刻度set(gca,'YLim',[1 26]);%X轴的数据显示范围
set(gca,'YTick',[0:1:26]);%设置要显示坐标刻度
% set(gca,'YTickLabel',[95:1:101]);%给坐标加标签 
grid on;%%
data =[input_label_test',output_label_test'];
data =data(:,1)-data(:,2);
AA=zeros(26,1)
for i=1:26number=0;for j=1:10if data( (i-1)*10+j ,1   )==0 number=number+1;end endAA(i,1)=number;
end%%confusion_matrix1( output_label_test,input_label_test)

　　FUNCTION  

confusion_matrix1 混淆矩阵画图

function confusion_matrix1(act1, det1)
%act1为真实的标签，det1为预测标签
%输入的是行向量
%https://blog.csdn.net/xuyingjie125/article/details/78417760
[mat, order] = confusionmat(act1, det1);
%confusionmat用来构建混淆矩阵，mat返回混淆矩阵，order返回每个变量所在的类
%mat(i,j)表示训练集中i类在测试集中被分到j类的个数
k = max(order);   %k为分类的个数imagesc(mat); %imagesc函数将矩阵mat的元素数值按大小转化为不同颜色，并在坐标轴对应位置以这种颜色染色
colormap(flipud(gray)); %转换成灰度图
%colormap函数是用来设定和获取当前色图的函数
%flipud是矩阵倒转命令，gray是matlab内置矩阵
%gray()会返回M×3的矩阵
title('混淆矩阵');
textStrings = num2str(mat(:),'%d');
%num2str(x，precision)把数组x转换成字符串形式表示，precision为精度
textStrings = strtrim(cellstr(textStrings));
%cellstr是将字符数组转换成cell类型
%裁切字符串的开头和尾部的空格，制表，回车符%meshgrid是MATLAB中用于生成网格采样点的函数
[x,y] = meshgrid(1:k);
hStrings = text(x(:),y(:),textStrings(:),'HorizontalAlignment','center');
%text函数表示对文字string定位于用坐标轴指定的位置，且对指定的属性进行设置
%HorizontalAlignment表示文字水平方向的对齐方式，center表示文本外框中间对齐
midValue = mean(get(gca,'CLim')); 
%get(gca,'CLim')获取当前的坐标轴句柄，并设置它的CLim(ColorLimit)值，用来使不同的图使用相同的色标
%mean函数返回均值，使得midValue获得图色标的均值0.5000
textColors = repmat(mat(:) > midValue,1,3);
%将矩阵[mat(:)>midValue]复制1×3块的矢量（颜色值必须为包含3个元素的数值矢量）
%把矩阵[mat(:)>midValue]作为矩阵textColors的元素
%B=repmat(A,M,N)这是一个处理大矩阵且内容有重复时使用，其功能是以A的内容堆叠在（M×N）的矩阵B中，
%B矩阵的大小由M×N及A矩阵的内容决定，
%如果A是一个3×4×5的矩阵，则矩阵B变为6×12×5set(hStrings,{'Color'},num2cell(textColors,2));
%n=1表示把每列的所有行转换为cell，n=2表示把每行的所有列转换为cell，将结构阵列转换成异质阵列
%然后set去重后放在hStrings%根据自己的分类要求更改类别标签set(gca,'XTick',1:26,'XTickLabel',{'1','2','3','4','5','6','7','8','9','10','11','12','13','14','15','16','17','18','19','20','21','22','23','24','25','26'},'YTick',1:26,'YTickLabel',{'1','2','3','4','5','6','7','8','9','10','11','12','13','14','15','16','17','18','19','20','21','22','23','24','25','26'});
% --------------------- 
% 浣滆?锛歴undreamoon 
% 鏉ユ簮锛欳SDN 
% 鍘熸枃锛歨ttps://blog.csdn.net/sundreamoon/article/details/79958207 
% 鐗堟潈澹版槑锛氭湰鏂囦负鍗氫富鍘熷垱鏂囩珷锛岃浆杞借闄勪笂鍗氭枃閾炬帴锛?

　　相关系数 code

%% 相关系数
%% V. 性能评价
%%
% 1. 相对误差error
input_test_label = actual_label;
error = abs(predict_label - input_test_label)./input_test_label;
[N,~]=size(input_test_label);
%%
% 2. 决定系数R^2
R2 = (N * sum(predict_label .* input_test_label) - sum(predict_label) * sum(input_test_label))^2 / ((N * sum((predict_label).^2) - (sum(predict_label))^2) * (N * sum((input_test_label).^2) - (sum(input_test_label))^2)); %%
% 3. 结果对比
result = [input_test_label' predict_label' error']%% VI. 绘图
figure
plot(1:N,input_test_label,'b:*',1:N,predict_label,'r-o')
legend('真实值','预测值')
xlabel('预测样本')
ylabel('标签值')
string = {'测试集预测结果对比';['R^2=' num2str(R2)]};
title(string)

　　SVM code

%% SVM神经网络的数据分类预测----意大利葡萄酒种类识别
% 
% 
% <html>
% <table border="0" width="600px" id="table1">	<tr>		<td><b><font size="2">该案例作者申明：</font></b></td>	</tr>	<tr>		<td><span class="comment"><font size="2">1：本人长期驻扎在此<a target="_blank" href="http://www.ilovematlab.cn/forum-158-1.html"><font color="#0000FF">板块</font></a>里，对<a target="_blank" href="http://www.ilovematlab.cn/thread-48362-1-1.html"><font color="#0000FF">该案例</font></a>提问，做到有问必答。</font></span></td></tr><tr>	<td><span class="comment"><font size="2">2：此案例有配套的教学视频，配套的完整可运行Matlab程序。</font></span></td>	</tr>	<tr>		<td><span class="comment"><font size="2">		3：以下内容为该案例的部分内容（约占该案例完整内容的1/10）。</font></span></td>	</tr>		<tr>		<td><span class="comment"><font size="2">		4：此案例为原创案例，转载请注明出处（<a target="_blank" href="http://www.ilovematlab.cn/">Matlab中文论坛</a>，<a target="_blank" href="http://www.ilovematlab.cn/forum-158-1.html">《Matlab神经网络30个案例分析》</a>）。</font></span></td>	</tr>		<tr>		<td><span class="comment"><font size="2">		5：若此案例碰巧与您的研究有关联，我们欢迎您提意见，要求等，我们考虑后可以加在案例里。</font></span></td>	</tr>		<tr>		<td><span class="comment"><font size="2">		6：您看到的以下内容为初稿，书籍的实际内容可能有少许出入，以书籍实际发行内容为准。</font></span></td>	</tr><tr>		<td><span class="comment"><font size="2">		7：此书其他常见问题、预定方式等，<a target="_blank" href="http://www.ilovematlab.cn/thread-47939-1-1.html">请点击这里</a>。</font></span></td>	</tr></table>
% </html>
% %% 清空环境变量
close all;
clear;
clc;
format compact;
%% 数据提取% 载入测试数据wine,其中包含的数据为classnumber = 3,wine:178*13的矩阵,wine_labes:178*1的列向量
load chapter12_wine.mat;% 画出测试数据的box可视化图
figure;
boxplot(wine,'orientation','horizontal','labels',categories);
title('wine数据的box可视化图','FontSize',12);
xlabel('属性值','FontSize',12);
grid on;% 画出测试数据的分维可视化图
figure
subplot(3,5,1);
hold on
for run = 1:178plot(run,wine_labels(run),'*');
end
xlabel('样本','FontSize',10);
ylabel('类别标签','FontSize',10);
title('class','FontSize',10);
for run = 2:14subplot(3,5,run);hold on;str = ['attrib ',num2str(run-1)];for i = 1:178plot(i,wine(i,run-1),'*');endxlabel('样本','FontSize',10);ylabel('属性值','FontSize',10);title(str,'FontSize',10);
end% 选定训练集和测试集% 将第一类的1-30,第二类的60-95,第三类的131-153做为训练集
train_wine = [wine(1:30,:);wine(60:95,:);wine(131:153,:)];
% 相应的训练集的标签也要分离出来
train_wine_labels = [wine_labels(1:30);wine_labels(60:95);wine_labels(131:153)];
% 将第一类的31-59,第二类的96-130,第三类的154-178做为测试集
test_wine = [wine(31:59,:);wine(96:130,:);wine(154:178,:)];
% 相应的测试集的标签也要分离出来
test_wine_labels = [wine_labels(31:59);wine_labels(96:130);wine_labels(154:178)];%% 数据预处理
% 数据预处理,将训练集和测试集归一化到[0,1]区间[mtrain,ntrain] = size(train_wine);
[mtest,ntest] = size(test_wine);dataset = [train_wine;test_wine];
% mapminmax为MATLAB自带的归一化函数
[dataset_scale,ps] = mapminmax(dataset',0,1);
dataset_scale = dataset_scale';train_wine = dataset_scale(1:mtrain,:);
test_wine = dataset_scale( (mtrain+1):(mtrain+mtest),: );
%% SVM网络训练
model = svmtrain(train_wine_labels, train_wine, '-c 2 -g 1');%% SVM网络预测
[predict_label, accuracy , pred] = svmpredict(test_wine_labels, test_wine, model);%% 结果分析% 测试集的实际分类和预测分类图
% 通过图可以看出只有一个测试样本是被错分的
figure;
hold on;
plot(test_wine_labels,'o');
plot(predict_label,'r*');
xlabel('测试集样本','FontSize',12);
ylabel('类别标签','FontSize',12);
legend('实际测试集分类','预测测试集分类');
title('测试集的实际分类和预测分类图','FontSize',12);
grid on;%%
% 
% <html>
% <table align="center" >	<tr>		<td align="center"><font size="2">版权所有：</font><a
% href="http://www.ilovematlab.cn/">Matlab中文论坛</a>   <script
% src="http://s3.cnzz.com/stat.php?id=971931&web_id=971931&show=pic" language="JavaScript" ></script> </td>	</tr></table>
% </html>
% 

　　 III  CNN

main

%%%  matlab实现LeNet-5
%%%  作者：xd.wp
%%%  时间：2016.10.22  14:29
%% 程序说明
%          1、池化（pooling）采用平均2*2
%          2、网络结点数说明：
%                           输入层：28*28
%                           第一层：24*24（卷积）*20
%                           tanh
%                           第二层：12*12（pooling）*20
%                           第三层：100(全连接)
%                           第四层：10(softmax)
%          3、网络训练部分采用800个样本，检验部分采用100个样本
clear all;clc;
%% 网络初始化
layer_c1_num=20;
layer_s1_num=20;
layer_f1_num=100;
layer_output_num=26;
%权值调整步进
yita=0.01;
%bias初始化
bias_c1=(2*rand(1,20)-ones(1,20))/sqrt(20);
bias_f1=(2*rand(1,100)-ones(1,100))/sqrt(20);
%卷积核初始化
[kernel_c1,kernel_f1]=init_kernel(layer_c1_num,layer_f1_num);
%pooling核初始化
pooling_a=ones(2,2)/4;
%全连接层的权值
weight_f1=(2*rand(20,100)-ones(20,100))/sqrt(20);
weight_output=(2*rand(100,layer_output_num)-ones(100,layer_output_num))/sqrt(100);
disp('网络初始化完成......');
%% 开始网络训练
disp('开始网络训练......');
for iter=1:10
for n=1:100for m=1:26%读取样本
%         train_data=imread(strcat(num2str(m),'_',num2str(n),'.bmp'));path = ['C:\Users\Administrator\Desktop\流程\时域频域图\figure_2\figure_28_28\',[num2str(m),'_',num2str(n)],'.bmp'];%如0_1~0_10train_data=imread(path);train_data=double(train_data);% 去均值
%       train_data=wipe_off_average(train_data);%前向传递,进入卷积层1for k=1:layer_c1_numstate_c1(:,:,k)=convolution(train_data,kernel_c1(:,:,k));%进入激励函数state_c1(:,:,k)=tanh(state_c1(:,:,k)+bias_c1(1,k));%进入pooling1state_s1(:,:,k)=pooling(state_c1(:,:,k),pooling_a);end%进入f1层[state_f1_pre,state_f1_temp]=convolution_f1(state_s1,kernel_f1,weight_f1);%进入激励函数for nn=1:layer_f1_numstate_f1(1,nn)=tanh(state_f1_pre(:,:,nn)+bias_f1(1,nn));end%进入softmax层for nn=1:layer_output_numoutput(1,nn)=exp(state_f1*weight_output(:,nn))/sum(exp(state_f1*weight_output));end%% 误差计算部分Error_cost=-output(1,m);
%         if (Error_cost<-0.98)
%             break;
%         end%% 参数调整部分[kernel_c1,kernel_f1,weight_f1,weight_output,bias_c1,bias_f1]=CNN_upweight(yita,Error_cost,m-1,train_data,...state_c1,state_s1,...state_f1,state_f1_temp,...output,...kernel_c1,kernel_f1,weight_f1,weight_output,bias_c1,bias_f1);end    
end
end
disp('网络训练完成，开始检验......');
%% 测试阶段
number = 1;  
count=0;
for n=101:110for m=1:26%读取样本
%         train_data=imread(strcat(num2str(m),'_',num2str(n),'.bmp'));path = ['C:\Users\Administrator\Desktop\流程\时域频域图\figure_2\figure_28_28\',[num2str(m),'_',num2str(n)],'.bmp'];%如0_1~0_10%读取样本train_data=imread(path);train_data=double(train_data);% 去均值
%       train_data=wipe_off_average(train_data);%前向传递,进入卷积层1for k=1:layer_c1_numstate_c1(:,:,k)=convolution(train_data,kernel_c1(:,:,k));%进入激励函数state_c1(:,:,k)=tanh(state_c1(:,:,k)+bias_c1(1,k));%进入pooling1state_s1(:,:,k)=pooling(state_c1(:,:,k),pooling_a);end%进入f1层[state_f1_pre,state_f1_temp]=convolution_f1(state_s1,kernel_f1,weight_f1);%进入激励函数for nn=1:layer_f1_numstate_f1(1,nn)=tanh(state_f1_pre(:,:,nn)+bias_f1(1,nn));end%进入softmax层for nn=1:layer_output_numoutput(1,nn)=exp(state_f1*weight_output(:,nn))/sum(exp(state_f1*weight_output));end[p,classify]=max(output);if (classify==m+1)count=count+1;endfprintf('真实类别为%d  预测类别为%d  概率值为%d \n',m,classify,p);infor(number,1) = m;infor(number,2) = classify;infor(number,3) = p;number = number+1;end
end%% 准确率计算number = 1;
A = size(infor);
for i=1:A(1)if infor(i,1)- infor(i,2)==0number =number+1;endendaccuracy = number/A(1)save data_1.mat

　FUNCTION  convolution

function [state]=convolution(data,kernel)%实现卷积层操作?
[data_row,data_col]=size(data);
[kernel_row,kernel_col]=size(kernel);
for m=1:data_col-kernel_col+1for n=1:data_row-kernel_row+1state(m,n)=sum(sum(data(m:m+kernel_row-1,n:n+kernel_col-1).*kernel));end
end

　　FUNCTION   convolution_f1

function [state_f1,state_f1_temp]=convolution_f1(state_s1,kernel_f1,weight_f1)
%% 完成卷积层2操作
layer_f1_num=size(weight_f1,2);
layer_s1_num=size(weight_f1,1);
%%
for n=1:layer_f1_numcount=0;for m=1:layer_s1_numtemp=state_s1(:,:,m)*weight_f1(m,n);count=count+temp;endstate_f1_temp(:,:,n)=count;state_f1(:,:,n)=convolution(state_f1_temp(:,:,n),kernel_f1(:,:,n));
end
end

　FUNCTION CNN_upweight

function [kernel_c1,kernel_f1,weight_f1,weight_output,bias_c1,bias_f1]=CNN_upweight(yita,Error_cost,classify,train_data,state_c1,state_s1,state_f1,state_f1_temp,...output,kernel_c1,kernel_f1,weight_f1,weight_output,bias_c1,bias_f1)
%%%     完成参数更新，权值和卷积核
%% 结点数目
layer_c1_num=size(state_c1,3);
layer_s1_num=size(state_s1,3);
layer_f1_num=size(state_f1,2);
layer_output_num=size(output,2);[c1_row,c1_col,~]=size(state_c1);
[s1_row,s1_col,~]=size(state_s1);[kernel_c1_row,kernel_c1_col]=size(kernel_c1(:,:,1));
[kernel_f1_row,kernel_f1_col]=size(kernel_f1(:,:,1));
%% 保存网络权值
kernel_c1_temp=kernel_c1;
kernel_f1_temp=kernel_f1;weight_f1_temp=weight_f1;
weight_output_temp=weight_output;
%% Error计算
label=zeros(1,layer_output_num);
label(1,classify+1)=1;
delta_layer_output=output-label;
%% 更新weight_output
for n=1:layer_output_numdelta_weight_output_temp(:,n)=delta_layer_output(1,n)*state_f1';
end
weight_output_temp=weight_output_temp-yita*delta_weight_output_temp;%% 更新bias_f1以及kernel_f1
for n=1:layer_f1_numcount=0;for m=1:layer_output_numcount=count+delta_layer_output(1,m)*weight_output(n,m);end%bias_f1delta_layer_f1(1,n)=count*(1-tanh(state_f1(1,n)).^2);delta_bias_f1(1,n)=delta_layer_f1(1,n);%kernel_f1delta_kernel_f1_temp(:,:,n)=delta_layer_f1(1,n)*state_f1_temp(:,:,n);
end
bias_f1=bias_f1-yita*delta_bias_f1;
kernel_f1_temp=kernel_f1_temp-yita*delta_kernel_f1_temp;
%% 更新weight_f1
for n=1:layer_f1_numdelta_layer_f1_temp(:,:,n)=delta_layer_f1(1,n)*kernel_f1(:,:,n);
end
for n=1:layer_s1_numfor m=1:layer_f1_numdelta_weight_f1_temp(n,m)=sum(sum(delta_layer_f1_temp(:,:,m).*state_s1(:,:,n)));end
end
weight_f1_temp=weight_f1_temp-yita*delta_weight_f1_temp;%% 更新 bias_c1
for n=1:layer_s1_numcount=0;for m=1:layer_f1_numcount=count+delta_layer_f1_temp(:,:,m)*weight_f1(n,m);   enddelta_layer_s1(:,:,n)=count;delta_layer_c1(:,:,n)=kron(delta_layer_s1(:,:,n),ones(2,2)/4).*(1-tanh(state_c1(:,:,n)).^2);delta_bias_c1(1,n)=sum(sum(delta_layer_c1(:,:,n)));
end
bias_c1=bias_c1-yita*delta_bias_c1;
%% 更新 kernel_c1
for n=1:layer_c1_numdelta_kernel_c1_temp(:,:,n)=rot90(conv2(train_data,rot90(delta_layer_c1(:,:,n),2),'valid'),2);
end
kernel_c1_temp=kernel_c1_temp-yita*delta_kernel_c1_temp;%% 网络权值更新
kernel_c1=kernel_c1_temp;
kernel_f1=kernel_f1_temp;weight_f1=weight_f1_temp;
weight_output=weight_output_temp;end

　　FUNCTION init_kernel

function [kernel_c1,kernel_f1]=init_kernel(layer_c1_num,layer_f1_num)
%卷积核初始化
for n=1:layer_c1_numkernel_c1(:,:,n)=(2*rand(5,5)-ones(5,5))/12;
end
for n=1:layer_f1_numkernel_f1(:,:,n)=(2*rand(12,12)-ones(12,12));
end
end

　　FUNCTION pooling

　

function state=pooling(data,pooling_a)
%% 实现取样层pooling操作
[data_row,data_col]=size(data);
[pooling_row,pooling_col]=size(pooling_a);
for m=1:data_col/pooling_colfor n=1:data_row/pooling_rowstate(m,n)=sum(sum(data(2*m-1:2*m,2*n-1:2*n).*pooling_a));end
end
end

　　view figure

clc
clear
load data_1.mat infor accuracy
%%
A = size(infor);
figure(1)
hold on
plot(infor(:,1),'bo');%实际类别
plot(infor(:,2),'r+');%predict_label
legend('实际测试集分类','预测测试集分类');
%设置x轴范围和刻度
set(gca,'XLim',[0 260]);%X轴的数据显示范围
set(gca,'XTick',[0:10:260]);%设置要显示坐标刻度
% set(gca,'XTickLabel',[0:1:10]);%给坐标加标签 %设置y轴范围和刻度set(gca,'YLim',[1 26]);%X轴的数据显示范围
set(gca,'YTick',[0:1:26]);%设置要显示坐标刻度
% set(gca,'YTickLabel',[95:1:101]);%给坐标加标签 
grid on;%%
data = transform(infor);
figure(2)
hold on
plot(data(:,1),'bo');%实际类别
plot(data(:,2),'r+');%predict_label
legend('实际测试集分类','预测测试集分类');
%设置x轴范围和刻度
set(gca,'XLim',[0 260]);%X轴的数据显示范围
set(gca,'XTick',[0:10:260]);%设置要显示坐标刻度
% set(gca,'XTickLabel',[0:1:10]);%给坐标加标签 %设置y轴范围和刻度set(gca,'YLim',[1 26]);%X轴的数据显示范围
set(gca,'YTick',[0:1:26]);%设置要显示坐标刻度
% set(gca,'YTickLabel',[95:1:101]);%给坐标加标签 
grid on;
%% 计算每个类别错误个数
error_cat_num = zeros(26,1);
for i=1:26number =0;for j= i*10-9 : i*10if data(j,1)- data(j,2)==0number =number+1;endenderror_cat_num(i,1)=number; end%%
%% 相关系数%%
% 1. 相对误差error
input_test_label = data(:,1);
predict_label = data(:,2);
error = abs(predict_label - input_test_label)./input_test_label;
[N,~]=size(input_test_label);
%%
% 2. 决定系数R^2
R2 = (N * sum(predict_label .* input_test_label) - sum(predict_label) * sum(input_test_label))^2 / ((N * sum((predict_label).^2) - (sum(predict_label))^2) * (N * sum((input_test_label).^2) - (sum(input_test_label))^2)); %%
% 3. 结果对比
result = [input_test_label' predict_label' error']%% VI. 绘图
figure
plot(1:N,input_test_label,'b:*',1:N,predict_label,'r-o')
legend('真实值','预测值')
xlabel('预测样本')
ylabel('标签值')
string = {'测试集预测结果对比';['R^2=' num2str(R2)]};
title(string)