Robust Multivariate Linear Regression Using the Student-t Distribution

Abstract

This brief report documents the Matlab function MVSREGRESS for multivariate linear regression based on the Student-t distribution. Student-t regression is more robust (less sensitive to occasional extreme observations) than normal (gaussian) linear regression. The report describes the statistical model and the ECM algorithm used to compute the maximum a-posteriori estimate of the model parameters. Two examples are presented to illustrate the robustness of Student regression compared to Normal regression.