CRoNoS & MDA 2019: Start Registration
View Submission - CRONOSMDA2019
A0266
Title: Using functions of inverse means and medians for dimension reduction Authors:  Andreas Artemiou - University of Limassol (Cyprus) [presenting]
Abstract: Two methods for dimension reduction in regression in the sufficient dimension reduction framework are introduced. The first method is using functions of sliced inverse means for dimension reduction using two different algorithmic approaches. One algorithm is computationally much faster than existing methodology. The other algorithm performs robustly in the case of discrete response variables (i.e. iris data), that is the ordering of the data does not affect the results. To robustify against outliers we present also a method that uses functions of inverse medians.