Qing Li
 

Department of Mathematics, Washington University
 

Title: Nonparametric Quasi-Likelihood and Its Application
 

Abstract: The quasi-likelihood method is a widely used approach particularly useful in generalized linear models (GLMs). Despite its versatility, it requires the specification of the variance function which is not always feasible. We will discuss the nonparametric quasi-likelihood (NQL) approach (Chiou and Muller, 1999) that only assumes the smoothness of the variance function. Main results of this method would be stated and, if time allows, we will talk about the application of this method on modeling microarray gene expression data.