目标识别笔记：鸢尾花数据集决策树/随机森林MATLAB仿真

clcclearload fisheriris %下载鸢尾花数据ctree &＃61; fitctree(meas,species); % meas是训练集&＃xff0c;species是真实结果view(ctree,&＃39;mode&＃39;,&＃39;graph&＃39;);X &＃61; meas;label &＃61; predict(ctree,X);count1 &＃61; 0; % 决策树决策结果for ii &＃61; 1:size(X,1)if ~strcmp(label(ii),species(ii))count1 &＃61; count1 &＃43; 1;endendfprintf(&＃39;决策树中共有%d个样本&＃xff0c;判断错误的有%d\n&＃39;,size(X,1),count1);Model &＃61; TreeBagger(50,meas,species);label &＃61; predict(Model,X);count2 &＃61; 0; % 随机森林决策结果 for ii &＃61; 1:size(X,1)if ~strcmp(label(ii),species(ii))count2 &＃61; count2 &＃43; 1;endendfprintf(&＃39;随机森林中共有%d个样本&＃xff0c;判断错误的有%d\n&＃39;,size(X,1),count2);

结果

决策树中共有150个样本&＃xff0c;判断错误的有3随机森林中共有150个样本&＃xff0c;判断错误的有0

决策树结果