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
Anna-Clara Monti, Universita degli Studi del Sannio Print
Friday, 26 November 2010, 14:30 - 15:30

Skew-t distribution: robust applications and inferential aspects

Anna-Clara Monti, Universita degli Studi del Sannio (Benevento)

Abstract : The skew-t distribution is a flexible model, which is able to cope with skewness and kurtosis in the data and, hence, it provides an alternative to the adoption of robust procedures, when departures from normality occur. Indeed the skew-t likelihood function can be used as robust likelihood in an adaptive estimation procedure. Furthermore, the skew-t distribution includes as special cases the Student t, the skew normal and the normal distribution. However, inference based on the skew t -model becomes problematic when the degrees of freedom diverge for two reasons: the expected information matrix is singular; and the parameter takes a value occurring at the boundary of its parameter space. The talk addresses these inferential concerns, considers alterative parameterizations and focus on accurate sub-model testing procedures to be implemented in the boundary-parameter case.

Location: 2NO salle des profs - Plaine
Contact: Jacqueline Bottemanne, This e-mail address is being protected from spam bots, you need JavaScript enabled to view it