知识卡片
训练算法与学得模型分开
训练算法从样本和设定中拟合模型,预测时把新输入送入已学得模型得到输出。新输入并不自动触发重训练;需要更新模型时还应明确标签、训练时机与版本,不能把每次调用都说成学习。
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
原书出处 · 1 处
- · 查看原书依据
ML models are built using training data. In regression models, the training set is a set of labeled examples [x, f(x)] where x is the input variables and f(x) is the label or observed target. The labels can be numerical or categorical. The goal of the ML algorithms is given a training set, find the approximation f of f(x) that best generalizes or predicts labels
参考来源
- 位置:《Machine Learning Fundamentals: Concepts, Models, and Applications》第03章《机器学习流程与数据管理》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Machine Learning Fundamentals: Concepts, Models, and Applications/_epub-src/OEBPS/xhtml/14_chapter03.xhtml)
- 结论依据:摘录中关于训练算法与学得模型分开的定义、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:ML models are built using training data. In regression models, the training set is a set of labeled examples [x, f(x)] where x is the input variables and f(x) is the label or observed target. The labels can be numerical or categorical. The goal of the ML algorithms is given a training set, find the approximation f of f(x) that best generalizes or predicts labels