← 返回刚才阅读的位置

知识卡片

普通读书笔记卡 · 9246 · 生成模型:GPT-6

增加预测列不保证样本外改善

理解等级:中级

在同一训练集的嵌套OLS模型中加变量,训练SSE不会增加,R²通常不下降,却可能不严格上升。更多变量可能使留出预测更差,应结合模型复杂度和独立评价,不按训练R²单独选模型。

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 coefficient of multiple determination is often used to check the adequacy of the regression model. The value of r2 lies between 0 and 1, or 0 and 100 percent, that is, 0 ≤ r2 ≤ 1. It indicates the fraction of total variation of the dependent variable y that is explained by the independent variables or predictors. Usually, the closer the value of R2

    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:摘录中关于增加预测列不保证样本外改善的定义、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 coefficient of multiple determination is often used to check the adequacy of the regression model. The value of r2 lies between 0 and 1, or 0 and 100 percent, that is, 0 ≤ r2 ≤ 1. It indicates the fraction of total variation of the dependent variable y that is explained by the independent variables or predictors. Usually, the closer the value of R2