Data Fusion Using Empirical Likelihood.docVIP

Open Journal of Statistics, 2012, 2, 547-556 /10.4236/ojs.2012.25070 Published Online December 2012 (http://www.SciRP.org/journal/ojs) Data Fusion Using Empirical Likelihood Hsiao-Hsuan Wang, Yuehua Wu, Yuejiao Fu, Xiaogang Wang Department of Mathematics and Statistics, York University, Toronto, Canada Email: stevenw@mathstat.yorku.ca Received September 28, 2012; revised October 30, 2012; accepted November 15, 2012 ABSTRACT The authors propose a robust semi-parametric empirical likelihood method to integrate all available information from multiple samples with a common center of measurements. Two different sets of estimating equations are used to im- prove the classical likelihood inference on the measurement center. The proposed method does not require the knowle- dge of the functional forms of the probability density functions of related populations. The advantages of the proposed method are demonstrated through extensive simulation studies by comparing the mean squared errors, coverage proba- bilities and average lengths of confidence intervals with those from the classical likelihood method. Simulation results suggest that our approach provides more informative and efficient inference than the conventional maximum likelihood estimator if certain structural relationship exists among the parameters of relevant samples. Keywords: Data Fusion; Empirical Likelihood; Robust Estimation; Multiple Samples 1. Introduction cond experiment since it comes from a different popu- lation unless a functional relationship between the two parameters is known. If the second population is related to the first one due to some connection between these two parameters, one should be able to utilize this con- nection and make better statistical inference. Different statistical methodologies have been proposed in the literature to integrate information from different sources (or populations) in a very general setting, see [1] and