知识卡片

普通读书笔记卡 · 9249.a · 生成模型:GPT-6

核超平面在特征空间仍是线性的

理解等级:中级

非线性核可让原空间边界弯曲,但在对应特征空间中分类超平面仍是线性的。核技巧不要求不断显式升维直到分开,也不保证任意冲突标签都能合理解决。

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

      This can be explained using another example. Suppose that the two sets of data (represented by solid and blank circles) are on a sheet and the sheet is lifted launching the points in the air. While the data points represented by balls are in the air, the sheet can be used to separate them. This lifting is mapping the data into a higher dimension and

    参考来源

    - 位置:《Machine Learning Fundamentals: Concepts, Models, and Applications》第06章《监督学习模型》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Machine Learning Fundamentals: Concepts, Models, and Applications/_epub-src/OEBPS/xhtml/19_chapter06.xhtml) - 结论依据:摘录中关于核超平面在特征空间仍是线性的的定义、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:This can be explained using another example. Suppose that the two sets of data (represented by solid and blank circles) are on a sheet and the sheet is lifted launching the points in the air. While the data points represented by balls are in the air, the sheet can be used to separate them. This lifting is mapping the data into a higher dimension and