知识卡片
多数类基准揭示准确率盲点
1000条中950条为非目标,恒预测非目标准确率为95%,原书95.5%是算术误差。这种分类器目标类召回为0,高总体准确率不证明能识别稀少目标,应同时检查目标类指标与代价。
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
- · 训练机器学习模型
原书出处 · 1 处
- · 查看原书依据
As discussed above, several metrics exist to assess the model performance of supervised machine learning models, and the choice of evaluation metric depends on the specific machine learning task being performed and the characteristics of the data. However, it is often good practice to use a combination of these metrics to perform model evaluation and select the most suitable model because one metric alone may
参考来源
- 位置:《Machine Learning Fundamentals: Concepts, Models, and Applications》第09章《训练机器学习模型》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Machine Learning Fundamentals: Concepts, Models, and Applications/_epub-src/OEBPS/xhtml/23_chapter09.xhtml)
- 结论依据:摘录中关于多数类基准揭示准确率盲点的定义、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:As discussed above, several metrics exist to assess the model performance of supervised machine learning models, and the choice of evaluation metric depends on the specific machine learning task being performed and the characteristics of the data. However, it is often good practice to use a combination of these metrics to perform model evaluation and select the most suitable model because one metric alone may