知识卡片

普通读书笔记卡 · 8739.b · 生成模型:GPT-6

业务指标映射模型指标须协商

理解等级:高级

模型指标常是业务结果的代理,错误类型可能影响不同目标,如误拦正常客户与漏掉欺诈各有代价。团队要先约定可接受平衡及基线,再优化模型,不能用总体准确率掩盖具体损失。

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

      Model measures Our ML models are a key outcome of our delivery efforts and potentially a key determinant of product performance. If we’re ambitious enough, with reinforcement learning for instance, we could allow our model or agent to directly optimize product or business measures such as acquisition. This approach can be de-risked by permitting only preapproved actions. For instance, product hero images might be selected

    参考来源

    - 位置:《Effective Machine Learning Teams (David Tan, Ada Leung, David Colls)》第02章《产品与交付实践》(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/ch02.html) - 结论依据:摘录中关于业务指标映射模型指标须协商的定义、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:Model measures Our ML models are a key outcome of our delivery efforts and potentially a key determinant of product performance. If we’re ambitious enough, with reinforcement learning for instance, we could allow our model or agent to directly optimize product or business measures such as acquisition. This approach can be de-risked by permitting only preapproved actions. For instance, product hero images might be selected