xlabel('样本','FontSize',10); ylabel('类别标签','FontSize',10); title('class','FontSize',10); for run = 2:14
subplot(3,5,run); hold on;
str = ['attrib ',num2str(run-1)]; for i = 1:178
plot(i,wine(i,run-1),'*'); end
xlabel('样本','FontSize',10); ylabel('属性值','FontSize',10); title(str,'FontSize',10); End
实验二：
%% SVM--葡萄酒种类识别
%机器学习作业 实验二：建立分类模型，进行类别标签预测 % 2015.6 lvyuxuan %
%% 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 lyx_wine.mat;
% 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
16
[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] = 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
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;
实验三：
%% SVM--葡萄酒种类识别
%机器学习作业 实验三：ROC图形分析 % 2015.6 lvyuxuan %
%% 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
17
阵,wine_labes:178*1的列向量 load lyx_wine.mat;
% 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
option = ['-t ' '2' ' -c ' '2' ' -g ' ' 1' ' -b ' '1'];
model = svmtrain(train_wine_labels, train_wine,option);
%% SVM网络预测
[predict_label,accuracy,dec_values] = svmpredict(test_wine_labels, test_wine, model,'-b 1'); %[predictlabel,accuracy,] = svmpredict(testlabel,test,model,'-b 1');
%% 结果分析
plotSVMroc(test_wine_labels,dec_values,3);
实验四：
%% SVM--葡萄酒种类识别
%机器学习作业 实验四：运用不同数量的训练样本及采用不同核函数的对比 % 2015.6 lvyuxuan %
18
%% 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 lyx_wine.mat;
% 选定训练集和测试集 x1=[10 15 20 25 30 35 40]; x2=[69 74 79 84 89 94 99];
x3=[140 145 150 155 160 165 170]; saccuracy1=[]; saccuracy2=[]; saccuracy3=[]; saccuracy4=[]; i=1;
%每类有i个训练数据 %每类有10个训练数据 for i=1:7
% 将第一类的1-10/15/20/25/30/35/40,第二类的60-69/74/79/84/89/94/99,第三类的131-140/145/150/155/160/165/170做为训练集
train_wine = [wine(1:x1(i),:);wine(60:x2(i),:);wine(131:x3(i),:)];
% 相应的训练集的标签也要分离出来
train_wine_labels = [wine_labels(1:x1(i));wine_labels(60:x2(i));wine_labels(131:x3(i))];
% 将第一类的11-59,第二类的80-130,第三类的141-178做为测试集 test_wine = [wine(x1(i)+1:59,:);wine(x2(i)+1:130,:);wine(x3(i)+1:178,:)];
% 相应的测试集的标签也要分离出来
test_wine_labels = [wine_labels(x1(i)+1:59);wine_labels(x2(i)+1:130);wine_labels(x3(i)+1:178)];
%% 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);
19
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
model1 = svmtrain(train_wine_labels, train_wine, '-c 2 -g 1 -t 0'); model2 = svmtrain(train_wine_labels, train_wine, '-c 2 -g 1 -t 1'); model3 = svmtrain(train_wine_labels, train_wine, '-c 2 -g 1 -t 2'); model4 = svmtrain(train_wine_labels, train_wine, '-c 2 -g 1 -t 3');
%% SVM网络预测
[predict_label, accuracy1] = svmpredict(test_wine_labels, test_wine, model1); saccuracy1=[saccuracy1 accuracy1];
[predict_label, accuracy2] = svmpredict(test_wine_labels, test_wine, model2); saccuracy2=[saccuracy2 accuracy2];
[predict_label, accuracy3] = svmpredict(test_wine_labels, test_wine, model3); saccuracy3=[saccuracy3 accuracy3];
[predict_label, accuracy4] = svmpredict(test_wine_labels, test_wine, model4); saccuracy4=[saccuracy4 accuracy4]; end
%% 结果分析
figure;
plot(x1,saccuracy1(1,:)','-^r','LineWidth',1,'MarkerFaceColor',[.49 1 .63],'MarkerSize',10);hold on; plot(x1,saccuracy2(1,:)','-oy','LineWidth',1,'MarkerFaceColor',[.49 1 .63],'MarkerSize',10);hold on;
plot(x1,saccuracy3(1,:)','-*b','LineWidth',1,'MarkerFaceColor',[.49 1 .63],'MarkerSize',10);hold on;
plot(x1,saccuracy4(1,:)','-sg','LineWidth',1,'MarkerFaceColor',[.49 1 .63],'MarkerSize',10);hold off; grid on;
title('采用不同核函数的对比')
legend('linear','polynomial','radial basis function','sigmoid'); xlabel('每类训练样本数量','FontSize',14);
ylabel('测试集预测分类准确率','FontSize',14);axis([10 45 75 100]);
20