← 返回刚才阅读的位置

知识卡片

普通读书笔记卡 · 9251.c · 生成模型:GPT-6

类别比较可省常数但概率需归一

理解等级:中级

同一输入X下,各候选类别共用证据P(X),求argmax可比较P(y)∏p(xᵢ|y)。输入换了证据通常也变,且该证据一般不是无条件边际的乘积;输出后验概率需按全部类别得分归一。

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 values for each probability can be obtained from the given data in Table 6.12. Substituting the values in the above equation will give the required probability in the above formula. For all entries in the data set, the denominator in the equation does not change, it remains static. Therefore, the denominator can be removed, and proportionality can be used. The equation can be written

    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:摘录中关于类别比较可省常数但概率需归一的定义、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 values for each probability can be obtained from the given data in Table 6.12. Substituting the values in the above equation will give the required probability in the above formula. For all entries in the data set, the denominator in the equation does not change, it remains static. Therefore, the denominator can be removed, and proportionality can be used. The equation can be written