知识卡片
信息增益比较加权子节点不纯度
分类树分裂应比较父节点不纯度与按子节点样本权重汇总的不纯度差。单看某个小子节点很纯可能忽略另一个混乱大节点;熵0表示类标签纯,并不自动表示样本外正确。
这张卡片看不懂?
这张卡片已通过人工审查,不再接受失败反馈。
已有反馈 · 0 条
还没有人提交反馈。
在整书笔记中的位置
- · 监督学习模型
原书出处 · 1 处
- · 查看原书依据
Information gain: it is a measure of the reduction in impurity achieved by splitting a data set on a particular feature in a decision tree. The splitting criterion is using the feature that offers the greatest information gain. Pruning: the process of removing branches from the tree that are redundant or do not provide any information or lead to overfitting. Figure 6.15A general structure decision
参考来源
- 位置:《Machine Learning Fundamentals: Concepts, Models, and Applications》第06章《监督学习模型》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Machine Learning Fundamentals: Concepts, Models, and Applications/_epub-src/OEBPS/xhtml/19_chapter06.xhtml)
- 结论依据:摘录中关于信息增益比较加权子节点不纯度的定义、比较或条件支撑卡片;原书表述不严谨的范围已在覆盖记录逐项限定,不把案例结果外推为普遍结论。
- 原始内容:Information gain: it is a measure of the reduction in impurity achieved by splitting a data set on a particular feature in a decision tree. The splitting criterion is using the feature that offers the greatest information gain. Pruning: the process of removing branches from the tree that are redundant or do not provide any information or lead to overfitting. Figure 6.15A general structure decision