分類預(yù)測 | MATLAB實現(xiàn)WOA-FS-SVM鯨魚算法同步優(yōu)化特征選擇結(jié)合支持向量機分類預(yù)測

效果一覽

基本介紹

MATLAB實現(xiàn)WOA-FS-SVM鯨魚算法同步優(yōu)化特征選擇結(jié)合支持向量機分類預(yù)測（完整程序和數(shù)據(jù)）
WOA鯨魚算法同步優(yōu)化特征選擇結(jié)合支持向量機分類預(yù)測，優(yōu)化前后對比，基于LIBSVM。

程序設(shè)計

• 完整程序和數(shù)據(jù)下載方式私信博主回復(fù)：MATLAB實現(xiàn)WOA-FS-SVM鯨魚算法同步優(yōu)化特征選擇結(jié)合支持向量機分類預(yù)測

%%  參數(shù)設(shè)置
% 定義優(yōu)化參數(shù)的個數(shù)，在該場景中，優(yōu)化參數(shù)的個數(shù)為數(shù)據(jù)集特征總數(shù) 。
%目標函數(shù)
fobj = @(x) fun(x,train_wine_labels,train_wine,test_wine_labels,test_wine); 
% 優(yōu)化參數(shù)的個數(shù) 特征維度
dim = size(train_wine,2); %特征維度
% 優(yōu)化參數(shù)的取值下限，[0,1],大于0.5為選擇該特征，小于0.5為不選擇該特征
lb = 0;
ub = 1;%%  參數(shù)設(shè)置
pop =10; %數(shù)量
Max_iteration=50;%最大迭代次數(shù)             
%% 優(yōu)化(這里主要調(diào)用函數(shù))
[Best_score,Best_pos,curve]=WOA(pop,Max_iteration,lb,ub,dim,fobj); 
figure
plot(curve,'linewidth',1.5);
xlabel('迭代次數(shù)');
ylabel('適應(yīng)度值');
title('收斂曲線');
grid on;c = 2;  
g = 2; 
toc
% 用優(yōu)化得到的特征進行訓(xùn)練和測試
cmd = ['-s 0 -t 2 ', '-c ', num2str(c), ' -g ', num2str(g), ' -q'];
model = libsvmtrain(train_wine_labels, train_wineNew, cmd);
test_wineNew = test_wine(:,B);
%% SVM網(wǎng)絡(luò)預(yù)測
[predict_labelTrain, accuracyTrain,~] = libsvmpredict(train_wine_labels, train_wineNew, model);
[predict_labelTest, accuracyTest,~] = libsvmpredict(test_wine_labels, test_wineNew, model);%% 基礎(chǔ)SVM預(yù)測結(jié)果
% 用優(yōu)化得到的特征進行訓(xùn)練和測試
cmd = ['-s 0 -t 2 ', '-c ', num2str(c), ' -g ', num2str(g), ' -q'];
model = libsvmtrain(train_wine_labels, train_wine, cmd);
%% SVM網(wǎng)絡(luò)預(yù)測
[predict_labelTrain1, accuracyTrain1,~] = libsvmpredict(train_wine_labels, train_wine, model);
[predict_labelTest1, accuracyTest1,~] = libsvmpredict(test_wine_labels, test_wine, model);%% 結(jié)果分析

參考資料

[1] https://blog.csdn.net/kjm13182345320/article/details/128163536?spm=1001.2014.3001.5502
[2] https://blog.csdn.net/kjm13182345320/article/details/128151206?spm=1001.2014.3001.5502