Menu Content/Inhalt
Seminars Print
previous year previous month next month next year
See by year See by month See by week See Today Search Jump to month
Alexis Decurninge, Huawei Technologies Print
Thursday, 08 October 2015, 12:15 - 13:15

Alexis Decurninge, Huawei Technologies

Multivariate L-Moments Based on Transports

Abstract: Univariate L-moments are expressed as projections of the quantile function onto an orthogonal basis of univariate polynomials. Their existence at any order if the expectation of the underlying random variable is finite makes them alternative to moments for the study of heavy-tailed distributions. Moreover, L-moment inference methods are attractive thanks to their simplicity since empirical L-moments are expressed in function of the order statistics.

We will present multivariate versions of L-moments expressed as collections of orthogonal projections of a multivariate quantile function on a basis of multivariate polynomials. We will consider quantile functions defined as transports from the uniform distribution on [0;1]^d onto the distribution of interest and present some properties of the subsequent L-moments. We will in particular detail the particular cases of Rosenblatt and optimal transports and present the estimators of these, involving tessellations (in the optimal transport case) and concomitants (in the Rosenblatt case).

Location: R42.2.110 !!!
Contact: Nancy De Munck, This e-mail address is being protected from spam bots, you need JavaScript enabled to view it