← 返回刚才阅读的位置

知识卡片

普通读书笔记卡 · 9242.c · 生成模型:GPT-6

异常筛查应覆盖上下两侧

理解等级:中级

上界μ+3σ不是数据的实际最大值,下界也不意味着下方一定没有记录。仅筛depth>max_val得到的是上侧候选数,不能把它直接称全部异常数量;还需实际检查低侧。

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

      From the above, the minimum value is 57.452 and the maximum value is 66.047. This indicates that there are no outliers at the lower tail of the distribution (see the histogram). The maximum value in the variable depth is 66.047. This indicates that there are values outside of the maximum values.

    参考来源

    - 位置:《Machine Learning Fundamentals: Concepts, Models, and Applications》第05章《使用数据》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Machine Learning Fundamentals: Concepts, Models, and Applications/_epub-src/OEBPS/xhtml/17_chapter05.xhtml) - 结论依据:摘录中关于异常筛查应覆盖上下两侧的定义、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:From the above, the minimum value is 57.452 and the maximum value is 66.047. This indicates that there are no outliers at the lower tail of the distribution (see the histogram). The maximum value in the variable depth is 66.047. This indicates that there are values outside of the maximum values.