CMStatistics 2016: Start Registration
View Submission - CMStatistics
B1544
Title: Maximum likelihood estimators under block maxima Authors:  Clement Dombry - Universite de Franche Comte (France)
Ana Ferreira - IST-ID (Portugal) [presenting]
Abstract: Two fundamental methods in Extreme Value Theory are the Block Maxima (BM) and the Peaks-Over-threshold (POT) and, two widely used methods of estimation for inferences in extremes when appliyng BM or POT are the maximum likelihood (ML) and probability weighted moment (PWM). We prove asymptotic normality of ML estimators under the BM approach and maximum domain of attraction conditions. Further, we compare its asymptotic performance with PWM estimators both under BM and POT approaches.