A data scientist builds an anomaly detection pipeline in SAS Viya. They want to flag observations whose Mahalanobis distance from the cluster center exceeds a threshold. Which SAS procedure or action set is most appropriate for computing Mahalanobis distances?
Select an answer to reveal the explanation.
Short Explanation and Infographic
Here's the deal — d is correct because both approaches work in SAS. PROC DISTANCE with METHOD=MAHAL computes Mahalanobis distances between observations.
Full explanation below image
Full Explanation
D is correct because both approaches work in SAS. PROC DISTANCE with METHOD=MAHAL computes Mahalanobis distances between observations. PROC IML can compute the quadratic form (x-mu)' Sigma^-1 (x-mu) explicitly using matrix operations for full control. B is partially relevant (PROC CORR produces the covariance matrix) but there is no MAHALANOBIS= function in a DATA step; it would require matrix algebra. Since both A and C are genuine SAS approaches, D is the most complete answer.