Abstract
This paper considers robust estimation of the seemingly unrelated regression equations (SURE) model. Following the infinitesimal approach to robustness, we characterize the class of optimal bounded-influence estimators and we propose two computationally simple estimators. We also present the results of a set of Monte Carlo experiments carried out to study the behavior of these estimators.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have