Wayne State University

Add An Event

Division of Research | Research Events

Warning Icon This event is in the past.
March 22, 2017 | 1:00 p.m. - 2:00 p.m.
Category: Seminar
Location: Faculty/Administration #1146 | Map
656 W. Kirby
Detroit, MI 48202
Cost: Free

Speaker:  Abhijit Mandal, Wayne State University

Title: Robust Tests Using the Density Power Divergence

In any parametric inference problem, the robustness of the procedure is a real concern. A procedure which retains a high degree of efficiency under the model and simultaneously provides stable inference under data contamination is preferable in any practical situation over another procedure which achieves its efficiency at the cost of robustness or vice versa. The density power divergence family of  Basu et al. (Biometrika, 1998) provides a flexible class of divergences where the adjustment between efficiency and robustness is controlled by a single parameter $beta$.  In this talk I will consider general tests of parametric hypotheses based on the density power divergence. The asymptotic null distribution and the robustness properties of the test statistic will be explored. The performance of the test will be explored through simulations and real data analysis.

A part of this talk will be based on the following papers:
1.  A. Basu, A. Mandal, N. Martin and L. Pardo (2013). Testing Statistical Hypotheses based on the Density Power Divergence. Annals of the Institute of Statistical Mathematics, Vol. 65, 319-348.
2. A. Ghosh, A. Mandal, N. Martin and L. Pardo (2016). Influence Analysis of Robust Wald-type Tests. Journal of Multivariate Analysis, Vol. 147, 102-126.


For more information about this event, please contact Department of Mathematics at 3135772479 or eb1208@wayne.edu.