Multivariate Location and Scatter Matrix Estimation Under Cellwise and Casewise Contamination

ANDY LEUNG; VÍCTOR J. YOHAI; RUBEN H. ZAMAR

COMPUTATIONAL STATISTICS AND DATA ANALYSIS

ELSEVIER SCIENCE BV

Lugar: Amsterdam; Año: 2017 vol. 111 p. 59 - 76

0167-9473

This paper considers the problem of multivariate location and scatter matrix estimation when the data contain cellwise and casewise outliers. A two-step approach was proposed to deal with this problem: First, apply a univariate filter to remove cellwise outliers and second, apply a generalized S-estimator to downweight casewise outliers.This paper improves this proposal in three main directions. First, a consistent bivariate filter is introduced to be used in combination with the univariate filter in therst step. Second, a new fast subsampling procedure is proposed to generate starting points for the generalized S-estimator in the second step. Third, a non-monotonic weight function for the generalized S-estimator is proposed to better deal with casewise outliers in high dimension. A simulation study and real data example show that,unlike the original two-step procedure, the modified two-step approach performs well for high dimension. Moreover, the modied procedure outperforms the original one and other state-of-the-art robust procedures under cellwise and casewise data contamination.