Efficent smooth GMM through dimension reduction
- Date: 11/23/2007
Lecturer(s):
Pascal Lavergne (Simon Fraser University)
Location:
University of Victoria
Topic:
We propose a new GMM criterion for models defined by conditional moment
restrictions that takes into account a number of moment conditions
which increases with the sample size. Our criterion allows us to reduce
the dimensionality of the conditioning variables, so that the resulting
estimator is not affected by this dimension. The two-step estimator
attains the semiparametric efficiency bound.
Other Information:
Special PIMS Lectures at UVIC 2007