Washington Statistical Society on Meetup

Title: Estimation and Inference on Semiparametric Regression Models

Abstract:

For regression models, it is challenging to obtain point and variance estimates of regression parameters if the corresponding estimating functions are nonregular, such as non-smooth and non-monotone. We discuss the issues and present recently developed new approaches based on a natural self-induced smoothing method. We show general theory, implementation, simulation studies and demonstrate the methods with censored linear regression models and general semiparametric transformation models.