← 返回刚才阅读的位置

知识卡片

普通读书笔记卡 · 9169.a · 生成模型:GPT-6

训练误差低不保证分布外泛化

理解等级:中级

优化只看已给样本,可能学到在训练数据上有效却在新场景失效的规则。更多样本若仍不代表部署条件,也不能自动解决泛化;需要独立评估以及对环境变化的具体覆盖。

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

      Third, the error between the prediction and the true class is only calculated based on the given training data. Even if we were to train with unlimited amounts of data points, it would still be difficult to generalize for new data and to know whether these new measurements will be distributed in a way that is representative of the training data.

    参考来源

    - 位置:《Introduction to Autonomous Robots Mechanisms, Sensors, Actuators, and Algorithms (Nikolaus Correll, Bradley Hayes etc.)》第10章《人工神经网络》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Introduction to Autonomous Robots Mechanisms, Sensors, Actuators, and Algorithms (Nikolaus Correll, Bradley Hayes etc.)/_epub-src/OEBPS/xhtml/chapter_10.xhtml) - 结论依据:摘录中关于训练误差低不保证分布外泛化的定义、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:Third, the error between the prediction and the true class is only calculated based on the given training data. Even if we were to train with unlimited amounts of data points, it would still be difficult to generalize for new data and to know whether these new measurements will be distributed in a way that is representative of the training data.