CMStatistics 2016: Start Registration
View Submission - CMStatistics
B1709
Title: Combining subsample estimators in quantile regression Authors:  Keith Knight - University of Toronto (Canada) [presenting]
Abstract: The aim is to explore the extent to which the classical regression quantile estimator can be ``improved'' by combining estimators from bootstrap samples or subsamples. Such estimators can have the same asymptotic efficiency as the classical estimator while being more robust to small changes in the data.