Multinomial Approximation and Sequential Simulation
2002/03/07
研討會日期 : 2002-03-07
時間 : 14:00
主講人 : Walter Beckert
地點 : B棟110室
演講者簡介 : Prof. Walter Beckert Ph.D. in Economics， U.C. Berkeley (2000)
現為Assistant Professor of Economics，University of Florida
其主要研究領域為Industrial Organization and Policy
演講摘要 : In many econometric problems of pratical interest, the classical Maximum Likelihood (ML) Theory breaks down, either because the ML estimator is analytically intractable or because it fails to exist. This talk describes an estimation methodology, built on a sequence of multinomial approximations and assisted by simulation if necessary, which yields consistent and efficient estimators in such situations. The properties of these estimators are discussed in the context of the Hajek - LeCam theory of Locally Asymptotically Quadratic (LAQ) families.
