STAT 6830 - Robust Statistical Analysis

Robust statistical procedures for inference in location, linear and multivariate models are presented. This will include broad classes of robust estimates, including R>-, M>- and L>-estimates of both regular and bounded influence types. Concepts such as breakdown point, influence function, and asymptotic theory are used to obtain properties of these procedures. Computational aspects of these estimates are discussed along with small sample properties and applications of these procedures.

Prerequisites/Corequisites: Prerequisites: STAT 6600 and 6620.

Credits 3 hrs.

Notes:
Open to Graduate Students Only.