[Statlist] Reminder *Research seminar in statistics* Université de Genève / Faculté SES

[Sylvie Huber]

RESEARCH SEMINAR IN STATISTICS



Organisers :                                                                                   .

E. Cantoni - E. Ronchetti - S. Sperlich - M-P. Victoria-Feser



Friday November 29th, 2013

at 11h15

Room M 5220, Uni Mail (40, bd du Pont-d'Arve)



Nicolai Meinshausen

ETH Zurich



"Nearly assumption-free confidence intervals for groups of variables in high-dimensional regression"

Abstract:

It is in general challenging to provide confidence intervals for individual variables in highdimensional regression without making strict or unverifiable assumptions on the design matrix. We show here that a confidence interval can be derived without making any assumptions on the design matrix. The lower bound for the regression coefficient of individual variables can be derived via linear programming. The idea also generalizes naturally to groups of variables, where we can derive a one-sided confidence interval for the joint e↵ect of a group. While the confidence intervals of individual variables are by the nature of the problem often very wide, it is shown to be possible to detect the contribution of groups of highly correlated predictor variables even when no individual variable shows a significant effect.



http://www.stat-center.unige.ch/ResSem.html



	[[alternative HTML version deleted]]