知识卡片
分布检查要联系数据生成
均值、分位数、Q-Q图和密度图可帮助发现偏斜、边界或异常值。评分本来有界、补零又可能产生堆积,拒绝正态假设不自动表示数据错误或模型不可用。应先检查这些形状如何形成,再决定处理方法。
这张卡片看不懂?
这张卡片已通过人工审查,不再接受失败反馈。
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 summary of these features is given in Table 2. Clearly, the mean is not equal to the median for both these features, so there is a chance that these might not follow a normal distribution. Also, note that there is a huge gap between the third quartile and the maximum value for TotalSales. On the other hand, the first and third quartiles generate a
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:摘录中关于分布检查要联系数据生成的定义、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 summary of these features is given in Table 2. Clearly, the mean is not equal to the median for both these features, so there is a chance that these might not follow a normal distribution. Also, note that there is a huge gap between the third quartile and the maximum value for TotalSales. On the other hand, the first and third quartiles generate a