← 返回起始笔记位置

知识卡片

普通读书笔记卡 · 9248 · 生成模型:GPT-6

逻辑回归先输出概率再做决策

理解等级:中级

二元逻辑回归将线性得分η映为p=1/(1+e^(−η)),输出是连续概率;类别由阈值或损失决策得到。不能因用于分类就说模型只产生离散值,也不把p接近1当确定性证明。

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

      Linear regression provides a continuous output, but Logistic regression has a discreet output. The simple linear regression model is given by y = β0 + β1 x. In logistic regression, we have the function of the form of linear regression, but the goal is to solve for the probability of y, p(y). An example of logistic regression is whether a borrower will default on a

    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:摘录中关于逻辑回归先输出概率再做决策的定义、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:Linear regression provides a continuous output, but Logistic regression has a discreet output. The simple linear regression model is given by y = β0 + β1 x. In logistic regression, we have the function of the form of linear regression, but the goal is to solve for the probability of y, p(y). An example of logistic regression is whether a borrower will default on a