知识卡片

普通读书笔记卡 · 6481.h.12 · 生成模型:GPT-6

离散化为关联分析改变问题

理解等级:中级

把评分或销量按分位数等阈值分成高低类别,能让规则关注类别共现,但会丢失类别内部的数值差异。阈值本身影响规则支持度和提升度,应保存分箱规则,不把离散后的关联当作原始连续关系的完整表示。

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

      rating is converted to a categorical variable, RatingType, with levels as Excellent, Super, Good, Bad, and Poor (high to low) as per heuristics and quartile value. Also, TotalSales is converted into a categorical variable, as discussed previously. Further note that each feature has several levels, and the associativity is actually calculated for each level. So, the overall associativity is taken as the median value found

    参考来源

    - 位置:《Artificial Intelligence and Data Science in Recommendation System: Current Trends, Technologies, and Applications》第12章《时尚零售评估与预测》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Artificial Intelligence and Data Science in Recommendation System: Current Trends, Technologies, and Applications/_epub-src/OEBPS/Text/9789815136746-C12.html) - 结论依据:摘录中关于离散化为关联分析改变问题的定义、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:rating is converted to a categorical variable, RatingType, with levels as Excellent, Super, Good, Bad, and Poor (high to low) as per heuristics and quartile value. Also, TotalSales is converted into a categorical variable, as discussed previously. Further note that each feature has several levels, and the associativity is actually calculated for each level. So, the overall associativity is taken as the median value found