We propose a mode regression estimator for the case in which the variate of interest is continuous and obsesrvable over its entire unbounded support. We show that the proposed mode regression estimator is consistent under mild conditions and has a tractable asymptotic distribution. An empirical illustration
is provided to highlight the practicality and usefulness of the estimator.
Presented by:
João Santos Silva, Department of Economics, University of Essex (joint with Gordon C.R. Kemp)
Date & time:
May 27, 2009 12:00 pm - May 27, 2009 1:00 pm
Internal seminars home