Abstract:
|
A testing and inference problem for the multidimensional semiparametric partial linear model is considered. First, we estimate the linear component. We also obtain a new test result for the linear component whose asymptotic power does not depend on the functional component. The proposed testing procedure is easy to implement numerically. We also study its numerical performance using simulated data. Finally, we also show that a functional component can be estimated adaptively over a wide range of functional spaces.
|