知识卡片
合成评价数据要验证代表性
没有权限将生产数据带入开发时,可使用合成数据探索和构造行为场景。生成器可能遗漏稀有关系或复制原偏差,应说明覆盖限制,不能因统计表面相近就宣称等同真实生产。
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
- · 模型自动化测试
原书出处 · 1 处
- · 查看原书依据
In some situations, we cannot have access to production data for testing in a nonproduction environment. In such scenarios, we can use tools such as Synthetic Data Vault and CheckList to generate production-like synthetic data for testing ML models and uncovering issues using test samples that are similar to production data in terms of distribution.
参考来源
- 位置:《Effective Machine Learning Teams (David Tan, Ada Leung, David Colls)》第06章《模型自动化测试》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Effective Machine Learning Teams (David Tan, Ada Leung, David Colls)/_epub-src/OEBPS/ch06.html)
- 结论依据:摘录中关于合成评价数据要验证代表性的定义、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:In some situations, we cannot have access to production data for testing in a nonproduction environment. In such scenarios, we can use tools such as Synthetic Data Vault and CheckList to generate production-like synthetic data for testing ML models and uncovering issues using test samples that are similar to production data in terms of distribution.