知识卡片
表示方案与模型一起验收
原始类别、筛选特征和独热编码改变了模型输入,比较时应在相同数据切分与评分规则下评估。某个表示对一种模型有效,不说明对全部模型同样有益;特征选择也应限制在训练信息内。
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
tested on three variations of the data, i.e., (i) Using the original data, which is referred to as Raw, (ii) Selecting important features which are tagged as a feature, and (iii) Using dummy variables for each of the categorical variables which are called as Dummy. The models are trained on 60% of the data, and the rest 40% is kept for comparison.
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:摘录中关于表示方案与模型一起验收的定义、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:tested on three variations of the data, i.e., (i) Using the original data, which is referred to as Raw, (ii) Selecting important features which are tagged as a feature, and (iii) Using dummy variables for each of the categorical variables which are called as Dummy. The models are trained on 60% of the data, and the rest 40% is kept for comparison.