Linear regression under model uncertainty.

This article explores the concept of model uncertainty in linear regression analysis and proposes a method to characterize mean and variance uncertainty in the response variable. The authors introduce a family of estimators for regression parameters and parameters related to model uncertainty, and establish their consistency. The approach is applied to three different applications, including a forecasting model for the S&P Index. The article provides a valuable contribution to the limited literature on regression analysis under model uncertainty. [Extracted from the article]