Decision Tree CrossValidation Error of .07 How to determine if acceptable??

Question

Justin 2014-10-2

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/157167-decision-tree-crossvalidation-error-of-07-how-to-determine-if-acceptable

评论： Justin 2014-10-9

taking this example of coding from documentation:

    leafs=logspace(1,2,10);
    N=numel(leafs);
    err=zeros(N,1);
    for n=1:N
        t=fitctree(x,y,'CrossVal','On','MinLeaf',leafs(n));
        err(n)=kfoldLoss(t);
    end
    plot(leafs,err)

the data I used in this resulted in a chart showing errors ranging from about .06 to .09.

Does anyone know how to interpret this error result?

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Bruno Pop-Stefanov 2014-10-7

1
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/157167-decision-tree-crossvalidation-error-of-07-how-to-determine-if-acceptable#answer_154275

Cross validation is a technique to assess how well a predictor can generalize to new data. The lower the cross validation error, the better. Having a CV error of .07 tells you that 93% of the test data set was correctly classified. Whether this is a good score depends on your application and your expectations. The point of this chart is to compare the performance of the classification tree for various leaf sizes. This chart will let you know what you should choose for the leaf size in order to get the best classification accuracy possible.