Model Averaging for Nonlinear Regression Models

This article considers the problem of model averaging for regression models that can be nonlinear in their parameters and variables. We consider a (NMA) framework propose weight-choosing criterion, information criterion (NIC). show up to constant, NIC is an asymptotically unbiased estimator risk function under settings with some mild assumptions. also prove optimality convergence weights. Monte Carlo experiments reveal NMA leads relatively lower risks compared alternative selection methods most situations. Finally, we apply method predicting individual wage, where our approach lowest prediction errors cases.