主讲人：Carlos Brunet Martins-Filho Professor
题目：Nonparametric Econometric Models
主讲人简介:Name: Carlos Brunet Martins-Filho
Personal: Born in Fortaleza, Brazil; Citizenship: Brazil and United States of America.
•Ph.D., Economics, University of Tennessee, USA, 1992.
• M.A., Economics, University of Tennessee, USA, 1991.
• B.S., Economics, Universidade Federal do Ceará, Brazil, 1988.
• Research interests: Econometrics, Statistics.
•August, 2009 - present: Professor, Department of Economics, University of Colorado at Boulder, USA.
• April, 2017 - present: Professor (affiliate), Department of Applied Mathematics, University of Coloradoat Boulder, USA.
• December, 2008 - present: Senior Research Fellow, International Food Policy Research Institute, Wash-ington D.C., USA.
讲座内容简介:Classical Econometric theory on inference and testing has been built for statistical models withfinite-dimensional parameter. This theory is very well developed and has proven extremely useful for empir-ical modeling inEconomics and other disciplines. However, in the last three decades,econometricians andstatisticians have come to recognize that it is often inadequate to assume that data generating processes ofinterest can be fully or partially described by a finite-dimensional parameter. This has lead to the emergenceof a growing literature on the specification and estimation of statistical models where the parameters ofinterest are functions or infinite dimensional vectors, commonly known as nonparametric models. In thisseminar, we present an overview of the most salient developments in the specification and estimation ofnonparametric models of density, distribution and regression. The presentation compares the two modelingapproaches, highlights differences in estimation and inference and discusses current deficiencies in the generaltheory of nonparametric estimation and inference.