10/18/2013

Post date: Oct 21, 2013 8:51:50 PM

Speaker: Dr. Chiung-Yu Huang, Division of Biostatistics and Bioinformatics, SKCCC, , JHU

Title: An Overview of Generalized Method of Moments and Empirical Likelihood Method

Abstract: Lars Hansen, a winner of the 2013 Nobel Prize in Economic Sciences, is best known for his work on Generalized Method of Moments (GMM). GMM provides an unified framework that includes many well-known estimators, such as least squares, maximum likelihood, and maximum pseudo likelihood estimators. In this talk, I will give a brief overview of the GMM approach (Hansen, 1982). I will also review the empirical likelihood (EL) method (Owen, 1988; Qin and Lawless, 1994) as an alternative estimation procedure to GMM. The two methods are asymptotically equivalent, yet the EL approach can provide better confidence intervals and tests with a moderate sample size. I will demonstrate the use of the two methods for combining auxiliary information in model estimation.