知识卡片
欠拟合与过拟合需分别诊断
欠拟合表现为训练数据本身也拟合不好,可能来自模型限制或优化不足;过拟合则是训练表现好而未见数据表现差。两种现象指向不同排查方向。测试表现差也可能来自分布变化或数据质量问题,不能把一切测试退化都简单归结参数太多。
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
- · 泛化问题
原书出处 · 1 处
- · 查看原书依据
A learner may perform poorly for one of two reasons: either it failed to optimize the objective on the training data, or it succeeded on the training data but in a way that does not generalize to the test setting. The former is called underfitting and the latter is called overfitting.
参考来源
- 位置:《Foundations of Computer Vision (Antonio Torralba, Phillip Isola etc.)》第11章《泛化问题》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Foundations of Computer Vision (Antonio Torralba, Phillip Isola etc.)/_epub-src/OPS/xhtml/c011.xhtml)
- 结论依据:摘录中关于欠拟合与过拟合需分别诊断的定义、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:A learner may perform poorly for one of two reasons: either it failed to optimize the objective on the training data, or it succeeded on the training data but in a way that does not generalize to the test setting. The former is called underfitting and the latter is called overfitting.