知识卡片
指标先区分评分与排序
评分预测检查数值是否接近真实评价,排序检查相关物品是否排到前面,用户与系统指标则观察体验和运行。不同任务可以共用模型,却不能不加说明地共用一个“准确率”;需先确定要支持哪种决定。
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 evaluation metrics are categorized as classification results, rating predictions, and ranking metrics. HitRatio@10 is a commonly used metric for RS. In general, it is a way to calculate the number of hits in an n-sized list of ranked items. The frequently used measures employed for RS have been discussed in this section and represented in Equations (1-9) [82].
参考来源
- 位置:《Artificial Intelligence and Data Science in Recommendation System: Current Trends, Technologies, and Applications》第03章《深度推荐与评估指标》(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-C3.html)
- 结论依据:摘录中关于指标先区分评分与排序的定义、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 evaluation metrics are categorized as classification results, rating predictions, and ranking metrics. HitRatio@10 is a commonly used metric for RS. In general, it is a way to calculate the number of hits in an n-sized list of ranked items. The frequently used measures employed for RS have been discussed in this section and represented in Equations (1-9) [82].