Martin Bilodeau is a statistician in the Département de mathématiques et de statistique at the Université de Montréal. His research covers robust multivariate estimation, S-estimators, nonparametric smoothing, and seemingly unrelated regression models. He is a co-developer of high breakdown-point S-estimators for SUR systems.
Key Contributions / Features
Bilodeau and Duchesne (2000) — "Robust Estimation of the SUR Model," The Canadian Journal of Statistics 28: 277–288: adapted S-estimators to the SUR model; proved affine equivariance and n-consistency under elliptical errors; modified Ruppert's (1992) algorithm for fast computation; bivariate diagnostic plots for multivariate outlier detection; Table 2 extends Lopuhaä (1989) asymptotic efficiency table to q equations.
Bilodeau (1992) — Fourier smoother and additive models, Canadian Journal of Statistics 20: 257–269: used in the Grunfeld example to confirm that nonparametric additive fits agree with Grunfeld's linear investment model.