Discrete Choice Methods with Simulation / Edition 2

by Kenneth E. Train

Overview

This book describes the new generation of discrete choice methods, focusing on the many advances that are made possible by simulation. Researchers use these statistical methods to examine the choices that consumers, households, firms, and other agents make. Each of the major models is covered: logit, generalized extreme value, or GEV (including nested and cross-nested logits), probit, and mixed logit, plus a variety of specifications that build on these basics. Simulation-assisted estimation procedures are investigated and compared, including maximum stimulated likelihood, method of simulated moments, and method of simulated scores. Procedures for drawing from densities are described, including variance reduction techniques such as anithetics and Halton draws. Recent advances in Bayesian procedures are explored, including the use of the Metropolis-Hastings algorithm and its variant Gibbs sampling. The second edition adds chapters on endogeneity and expectation-maximization (EM) algorithms. No other book incorporates all these fields, which have arisen in the past 25 years. The procedures are applicable in many fields, including energy, transportation, environmental studies, health, labor, and marketing.

Product Details

ISBN-13:	9780521747387
Publisher:	Cambridge University Press
Publication date:	06/30/2009
Edition description:	New Edition
Pages:	400
Product dimensions:	6.00(w) x 8.90(h) x 0.90(d)

Table of Contents

1. Introduction; Part I. Behavioral Models: 2. Properties; 3. Logit; 4. GEV; 5. Probit; 6. Mixed logit; 7. Variations on a theme; Part II. Estimation: 8. Numerical maximization; 9. Drawing from densities; 10. Simulation-assisted estimation; 11. Individual-level parameters; 12. Bayesian procedures; 13. Endogeneity; 14. EM algorithms.

Editorial Reviews

"Simulation methods have unshackled discrete choice analysis, breaking down the computational barriers to use of plausible, interpretable models. Kenneth Train provides an excellent road map for both econometric specialists and practitioners through this comprehensive, readable treatment that pulls together the research literature and provides many new and useful results." Daniel McFadden, Nobel Laureate, University of California, Berkeley

"An outstanding textbook for advanced students and a reference for experienced practitioners of discrete choice analysis. The text covers modern simulation methods that advanced choice modelers should know. The book is blessed by Kenneth Train's unique gift for simplifying and explaining the topic." Moshe Ben-Akiva, Massachusetts Institute of Technology

"A must have, must read book for academics and practitioners interested in understanding, modelling and predicting decision-making and choice behavior. As we have come to expect from Ken, his new book makes very complex topics accessible to a wide audience. The book covers the basics through to leading-edge thought and work in complex model systems using Bayesian and simulation estimation methods. But wait there's more! There's ‘Ken the movie'! You also can watch and listen to Ken's lectures on each topic via his UC Berkeley website and access his lecture notes. An unbeatable package for serious students that sets a new standard in educational communication for the field of probabilistic discrete choice modelling." Jordan Louviere, University of Technology, Sydney, Australia

"Ken Train's book provides outstanding coverage of the most advanced elements of the estimation and usage of discrete choice models that require simulation to take account of randomness in the population under study. His writing is clear and understandable providing both the new and experienced reader with excellent insights into and understanding of all aspects of these new and increasingly important methods." Frank S. Koppelman, Northwestern University

"This is a masterful book, authored by one of the leading contributors to discrete choice methods and analysis. No other book covers this ground with such up-to-date detail in respect of theory and implementation. The chapters on simulation and recent developments such as mixed logit are most lucid. As a text or reference work this volume should have currency for a long time. It will appeal to the practitioner as much as to the specialist researcher who has been in this field for many years." David Hensher, The University of Sydney

"Simulation based estimation is a major advance in econometrics and discrete choice modeling. The technique has revolutionized both classical and Bayesian analysis. Ken Train's many papers have made a large contribution to this literature. Discrete Choice Methods with Simulation collects these results in a comprehensive, up-to-date source, with chapters on behavioral foundations, theoretical and practical aspects of estimation and a variety of applications. This book is a thoroughly enjoyable blend of theory, analysis, and case studies; it is a complete reference for developers and practitioners." William Greene, New York University

"The author's writing is clear and lucid. He succeeds in providing the reader with excellent insights and understanding of the methodology. This book is an excellent blend of theory, analysis, and real problems. Because of this, I would rate it as an exceptionally valuable resource." Journal of the American Statistical Association, Karla V. Ballman, Mayo Clinic College of Medicine

From the Publisher