知识卡片
距离模型对特征尺度敏感
欧氏距离中大尺度特征可能主导邻近判断,缩放可调整各维的影响。尺度选择本身是一项建模判断,不保证所有模型都必须标准化或变换后必更准确;目标缩放还要说明如何还原输出单位。
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
- · 使用数据
原书出处 · 1 处
- · 查看原书依据
These differences in the ranges of initial features cause trouble for many machine learning models. For example, suppose a model is based on distance computation, if one of the features has a large range or variation, the distance will be governed by this particular feature. Using data standardization, the data are converted to a standard format that is easily understood and processed by computers.
参考来源
- 位置:《Machine Learning Fundamentals: Concepts, Models, and Applications》第05章《使用数据》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Machine Learning Fundamentals: Concepts, Models, and Applications/_epub-src/OEBPS/xhtml/17_chapter05.xhtml)
- 结论依据:摘录中关于距离模型对特征尺度敏感的定义、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:These differences in the ranges of initial features cause trouble for many machine learning models. For example, suppose a model is based on distance computation, if one of the features has a large range or variation, the distance will be governed by this particular feature. Using data standardization, the data are converted to a standard format that is easily understood and processed by computers.