知识卡片
Gini越小表示节点越纯
Gini=1−Σpₖ²,纯节点为0,K类均匀时最大为1−1/K,二类最大0.5。原文说靠近1更理想把方向弄反,二元分裂也不表示只能处理二类目标。
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
Gini impurity is a metric that evaluates the accuracy of a split among the classified groups. It is an evaluation metric used to evaluate the decision tree model. It stores the sum of the squared probabilities of each class. It computes the degree of probability of a specific variable that is wrongly being classified. Its value is between 0 and 1. A value of 0
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Gini越小表示节点越纯的定义、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:Gini impurity is a metric that evaluates the accuracy of a split among the classified groups. It is an evaluation metric used to evaluate the decision tree model. It stores the sum of the squared probabilities of each class. It computes the degree of probability of a specific variable that is wrongly being classified. Its value is between 0 and 1. A value of 0
- 生成时纠错核验:[官方机制定义](https://scikit-learn.org/stable/modules/tree.html#classification-criteria)(2026-11-06;非代码执行或独立内容审查)。