知识卡片

普通读书笔记卡 · 9168 · 生成模型:GPT-6

分类与回归先确定输出语义

理解等级:中级

分类预测离散类别或其分布,回归预测连续数值或向量。把距离离散分箱可转成分类任务,但分箱会改变误差和分辨率要求;不能只改网络输出数量就认为两种任务评价标准相同。

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

      Remember: A classification problem requires that input data be classified between two or more classes; a regression problem requires a prediction of a (possibly continuous or high-dimensional) quantity. While a regression problem can be converted into a classification one (and vice versa), the machine learning community generally considers them two separate applications, and different techniques are developed to perform in each of these domains.

    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:摘录中关于分类与回归先确定输出语义的定义、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:Remember: A classification problem requires that input data be classified between two or more classes; a regression problem requires a prediction of a (possibly continuous or high-dimensional) quantity. While a regression problem can be converted into a classification one (and vice versa), the machine learning community generally considers them two separate applications, and different techniques are developed to perform in each of these domains.