知识卡片

普通读书笔记卡 · 9251 · 生成模型:GPT-6

森林特征随机性作用于分裂搜索

理解等级:中级

随机森林在重抽样基础上,对每个节点的候选分裂特征做随机限制以减少树相关性。它不只是给每棵树永久固定一个特征子集,也不保证树彼此独立或超参数只有三项。

这张卡片看不懂?

已有反馈 · 0 条

    还没有人提交反馈。

    在整书笔记中的位置

    原书出处 · 1 处

    • · 查看原书依据

      The random forest algorithm is an extension of the bagging method. It uses both bagging and feature randomness to create a forest of decision trees that are uncorrelated. Feature randomness or feature bagging generates a random subset of features with low correlation among the decision trees. The main difference between the decision trees and random forests is that the decision trees consider all possible feature

    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:摘录中关于森林特征随机性作用于分裂搜索的定义、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:The random forest algorithm is an extension of the bagging method. It uses both bagging and feature randomness to create a forest of decision trees that are uncorrelated. Feature randomness or feature bagging generates a random subset of features with low correlation among the decision trees. The main difference between the decision trees and random forests is that the decision trees consider all possible feature