SVM.rar_SVM matlab_SVM MATLAB分类_classification_matlab svm_svm +

%% SVM分类预测----炉温趋势种类识别 %% 清空环境变量 close all; clear; clc; format compact; %% 数据提取 % 载入测试数据 % 类别标签classnumber = 3,样本输入two:90*5的矩阵（风量、当前炉与上一炉风量差值、热风压力差值、压差差值、透气性差值）,样本的类别输出label:90*1的列向量 load classnumber.mat; load label.mat; load two.mat; % 画出测试数据的分维可视化图 figure subplot(3,2,1); hold on for run = 1:90 plot(run,label(run),'*'); end xlabel('样本','FontSize',10); ylabel('类别标签','FontSize',10); title('class','FontSize',10); for run = 2:6 subplot(3,2,run); hold on; str = ['attrib ',num2str(run-1)]; for i = 1:90 plot(i,two(i,run-1),'*'); end xlabel('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1-10,第二类的31-40,第三类的61-70做为训练集 train_temper = [two(1:10,:);two(31:40,:);two(61:70,:)]; % 相应的训练集的标签也要分离出来 train_temper_label = [label(1:10,:);label(31:40,:);label(61:70,:)]; % 将第一类的11-30,第二类的41-60,第三类的71-90做为测试集 test_temper = [two(11:30,:);two(41:60,:);two(71:90,:)]; % 相应的测试集的标签也要分离出来 test_temper_label = [label(11:30,:);label(41:60,:);label(71:90,:)]; %% 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_temper); [mtest,ntest] = size(test_temper); dataset = [train_temper;test_temper]; % mapminmax为MATLAB自带的归一化函数 [dataset_scale,ps] = mapminmax(dataset',0,1); dataset_scale = dataset_scale'; train_temper = dataset_scale(1:mtrain,:); test_temper = dataset_scale( (mtrain+1):(mtrain+mtest),: ); %% SVM网络训练 model = svmtrain(train_temper_label, train_temper, '-s 0 -t 0 -c 2 -g 1'); %% SVM网络预测 [predict_label, accuracy] = svmpredict(test_temper_label, test_temper, model); %% 结果分析 % 测试集的实际分类和预测分类图 figure; hold on; plot(test_temper_label,'o'); plot(predict_label,'r*'); xlabel('测试样本集','FontSize',12); ylabel('类别标签','FontSize',12); legend('样本实际分类','样本预测分类'); grid on;