You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
------------------ 原始邮件 ------------------
发件人: ***@***.***>;
发送时间: 2022年4月21日(星期四) 下午5:24
收件人: ***@***.***>;
抄送: ***@***.***>; ***@***.***>;
主题: Re: [zhaoxingfeng/RandomForest] 模型的min_sample_leaf问题 (Issue #12)
你的考虑是对的,代码中的<=改为<更合适,这样更符合叶子结点最小样本数的定义。
—
Reply to this email directly, view it on GitHub, or unsubscribe.
You are receiving this because you authored the thread.Message ID: ***@***.***>
您好老师,请问一下,这样在分裂的时候,当子节点样本数小于等于min_sample_leaf时,父节点停止分裂,这样感觉min_sample_leaf的设置有点奇怪,比如min_sample_leaf=1,就是他可能会在子节点中划分到1和xx,就删除了子节点,只划分到父节点了,这样会导致子节点中的最小样本数为2,我在pima数据集上做,和sklearn上实现的决策树相比,他的节点数少了一半以上,那这样的设置,是不是在上图应该把小于等于改成小于。
有点疑问
期待您的回复
The text was updated successfully, but these errors were encountered: