Muutke küpsiste eelistusi

Discrete Choice Methods with Simulation 2nd Revised edition [Kõva köide]

4.60/5 (43 hinnangut Goodreads-ist)
(University of California, Berkeley)
  • Formaat: Hardback, 400 pages, kõrgus x laius x paksus: 229x152x25 mm, kaal: 760 g, 17 Tables, unspecified; 46 Line drawings, unspecified
  • Ilmumisaeg: 06-Jul-2009
  • Kirjastus: Cambridge University Press
  • ISBN-10: 0521766559
  • ISBN-13: 9780521766555
Teised raamatud teemal:
Discrete Choice Methods with Simulation 2nd Revised editionSuurem pilt
  • Formaat: Hardback, 400 pages, kõrgus x laius x paksus: 229x152x25 mm, kaal: 760 g, 17 Tables, unspecified; 46 Line drawings, unspecified
  • Ilmumisaeg: 06-Jul-2009
  • Kirjastus: Cambridge University Press
  • ISBN-10: 0521766559
  • ISBN-13: 9780521766555
Teised raamatud teemal:
Püsilink: https://www.kriso.ee/db/9780521766555.html
Märksõnad:
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.

Muu info

This book describes the new generation of discrete choice methods, focusing on the many advances that are made possible by simulation.
1 Introduction 1
1.1 Motivation
1
1.2 Choice Probabilities and Integration
3
1.3 Outline of Book
7
1.4 A Couple of Notes
8
Part I Behavioral Models
2 Properties of Discrete Choice Models
11
2.1 Overview
11
2.2 The Choice Set
11
2.3 Derivation of Choice Probabilities
14
2.4 Specific Models
17
2.5 Identification of Choice Models
19
2.6 Aggregation
29
2.7 Forecasting
32
2.8 Recalibration of Constants
33
3 Logit
34
3.1 Choice Probabilities
34
3.2 The Scale Parameter
40
3.3 Power and Limitations of Logit
42
3.4 Nonlinear Representative Utility
52
3.5 Consumer Surplus
55
3.6 Derivatives and Elasticities
57
3.7 Estimation
60
3.8 Goodness of Fit and Hypothesis Testing
67
3.9 Case Study: Forecasting for a New Transit System
71
3.10 Derivation of Logit Probabilities
74
4 GEV
76
4.1 Introduction
76
4.2 Nested Logit
77
4.3 Three-Level Nested Logit
86
4.4 Overlapping Nests
89
4.5 Heteroskedastic Logit
92
4.6 The GEV Family
93
5 Probit
97
5.1 Choice Probabilities
97
5.2 Identification
100
5.3 Taste Variation
106
5.4 Substitution Patterns and Failure of IIA
108
5.5 Panel Data
110
5.6 Simulation of the Choice Probabilities
114
6 Mixed Logit
134
6.1 Choice Probabilities
134
6.2 Random Coefficients
137
6.3 Error Components
139
6.4 Substitution Patterns
141
6.5 Approximation to Any Random Utility Model
141
6.6 Simulation
144
6.7 Panel Data
145
6.8 Case Study
147
7 Variations on a Theme
151
7.1 Introduction
151
7.2 Stated-Preference and Revealed-Preference Data
152
7.3 Ranked Data
156
7.4 Ordered Responses
159
7.5 Contingent Valuation
164
7.6 Mixed Models
166
7.7 Dynamic Optimization
169
Part II Estimation
8 Numerical Maximization
185
8.1 Motivation
185
8.2 Notation
185
8.3 Algorithms
187
8.4 Convergence Criterion
198
8.5 Local versus Global Maximum
199
8.6 Variance of the Estimates
200
8.7 Information Identity
202
9 Drawing from Densities
205
9.1 Introduction
205
9.2 Random Draws
205
9.3 Variance Reduction
214
10 Simulation-Assisted Estimation
237
10.1 Motivation
237
10.2 Definition of Estimators
238
10.3 The Central Limit Theorem
245
10.4 Properties of Traditional Estimators
247
10.5 Properties of Simulation-Based Estimators
250
10.6 Numerical Solution
257
11 Individual-Level Parameters
259
11.1 Introduction
259
11.2 Derivation of Conditional Distribution
262
11.3 Implications of Estimation of 9
264
11.4 Monte Carlo Illustration
267
11.5 Average Conditional Distribution
269
11.6 Case Study: Choice of Energy Supplier
270
11.7 Discussion
280
12 Bayesian Procedures
282
12.1 Introduction
282
12.2 Overview of Bayesian Concepts
284
12.3 Simulation of the Posterior Mean
291
12.4 Drawing from the Posterior
293
12.5 Posteriors for the Mean and Variance of a Normal Distribution
294
12.6 Hierarchical Bayes for Mixed Logit
299
12.7 Case Study: Choice of Energy Supplier
305
12.8 Bayesian Procedures for Probit Models
313
13 Endogeneity
315
13.1 Overview
315
13.2 The BLP Approach
318
13.3 Supply Side
328
13.4 Control Functions
334
13.5 Maximum Likelihood Approach
340
13.6 Case Study: Consumers' Choice among New Vehicles
342
14 EM Algorithms
347
14.1 Introduction
347
14.2 General Procedure
348
14.3 Examples of EM Algorithms
355
14.4 Case Study: Demand for Hydrogen Cars
365
Bibliography 371
Index 385
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.