Title: Least squares estimation for nonlinear regression models with heteroscedasticity
Authors: Qiying Wang - University of Sydney (Australia) [presenting]
Abstract: Asymptotic theory is developed for general nonlinear regression models, establishing a new framework on least squares estimation that is easy to apply for various nonlinear regression models with heteroscedasticity. An application of the framework to nonlinear regression models with nonstationarity and heteroscedasticity is explored. Accompanying with these main results, a maximum inequality for a class of martingales is provided and some new results are established on convergence to a local time and convergence to a mixture of normal distributions.