Mahalanobis距離是一個可以準確找出資料分布上面極端值(Outliers)的統計方法,使用線性迴歸的概念,也就是說他使用的是共變數矩陣以及該資料分布的平均數來找尋極端值的產生,而可以讓一群資料系統具有穩健性(Robust),去除不必要的雜訊訊息,這邊拿前面共變數矩陣的資料為例,並且新增了兩個點座標向量來做Mahalanobis距離的比較-Mahalanobis distance is the information that can accurately identify the distribution of the above extreme values (Outliers) statistical methods, the concept of using a linear regression, meaning that he used the covariance matrix and the distribution of the information to find the extreme values of the average production of , but can allow a group of information systems with the robustness (Robust), removing unnecessary noise message here get in front of the data covariance matrix, for example, and added two points, coordinates of vectors to do comparison of Mahalanobis distance