Irène Gijbels – Department of Mathematics, University of Leuven, Belgium
Abstract: In univariate settings the notion of measuring skewness of a distribution is well
established. Extending this to measuring skewness (or asymmetry) in a multivariate
setting is less obvious. In this talk we start by briefly reviewing some notions of
asymmetry measures. We next propose a novel functional asymmetry measure (index)
which is based on the natural idea of reflective symmetry around the mode. The
proposed index is extended to the multivariate setting and a summarizing scalar (or
vector based index in multivariate settings) is derived from it. Some illustrative
examples are provided, and estimation of the proposed asymmetry index is discussed
in parametric and nonparametric (multivariate) settings.
Martedì 4 marzo 2025 ore 14.30
Area di Ricerca Napoli1
Via Pietro Castellino, 111, Napoli
Aula Conferenze Vaccaro