标题：Model Averaging based on Kullback-Leibler Distance

汇报人：Xinyu Zhang
，Academy of Mathematics and Systems Science, Chinese Academy of Sciences

功夫：2013年6月19日（周三）15:00-16:00

地址：拉斯维加斯98882号楼217室

Abstract:

Model averaging is an alternative to model selection for dealing with model uncertainty. This paper proposes a model averaging method based on Kullback-Leibler distance. The resulting model average estimator is proved to be asymptotically optimal. When combining least squares estimators, the model average estimator is proven to have the same large sample properties as the Mallows model average (MMA) estimator developed by Hansen (2007). We show via simulations that, in terms of mean square prediction error and mean square parameter estimation error,

in smaller sample size situations the proposed model average estimator is more efficient than the MMA estimator and the estimator based on model selection using the corrected Akaike information criterion. A modified version of the new model average estimator is further suggested for the case of heteroscedastic random errors. The method is further applied to a data set from Hong Kong real estate market.