← 返回起始笔记位置

知识卡片

普通读书笔记卡 · 8762.a · 生成模型:GPT-6

群体分差只是公平调查线索

理解等级:中级

分组指标可发现潜在不均衡,却不能单凭一组分数低就证明歧视或确定成因。应结合样本、真实标签质量、决策后果与适用公平标准调查;是否能采集敏感属性也需具体合法用途判断。

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

      ethical bias testing, to test if any model is inadvertently disadvantageous or systematically erroneous for certain segments of the population. For example, we could enumerate dimensions of potential sociodemographic bias and harm (e.g., race, gender, class, etc.) and test for any potential issues in each data segment. While we often can’t use these sociodemographic features to train our models, we can use them to uncover

    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:摘录中关于群体分差只是公平调查线索的定义、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:ethical bias testing, to test if any model is inadvertently disadvantageous or systematically erroneous for certain segments of the population. For example, we could enumerate dimensions of potential sociodemographic bias and harm (e.g., race, gender, class, etc.) and test for any potential issues in each data segment. While we often can’t use these sociodemographic features to train our models, we can use them to uncover