知识卡片
离群值会影响平方残差拟合
若数据混有错误回波或多种结构,单个最小二乘模型会折中照顾大残差,可能偏离真正目标。应先考虑模型是否正确、分段或异常识别,而不是把较小总平方误差当作所有点都来自同一条墙的证明。
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
- · 特征提取
原书出处 · 1 处
- · 查看原书依据
If the number of outliers is large, a least-squares fit will generate poor results because it will be the “best” fit that accomodates both inliers and outliers. Note that this generally results in a very poor fit, since the least-squares fit mathematically assigns a significant weight to outliers—not treating them as a single measurement to be discarded but a cumulative error to be reduced in
参考来源
- 位置:《Introduction to Autonomous Robots Mechanisms, Sensors, Actuators, and Algorithms (Nikolaus Correll, Bradley Hayes etc.)》第09章《特征提取》(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_9.xhtml)
- 结论依据:摘录中关于离群值会影响平方残差拟合的定义、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:If the number of outliers is large, a least-squares fit will generate poor results because it will be the “best” fit that accomodates both inliers and outliers. Note that this generally results in a very poor fit, since the least-squares fit mathematically assigns a significant weight to outliers—not treating them as a single measurement to be discarded but a cumulative error to be reduced in