A Self-Instructing Course in Mode Choice Modeling
Foreword
Original Acknowledgements
R Packages
1
Introduction
1.1
Background
1.2
Use of Disaggregate Discrete Choice Models
1.3
Application Context in Current Course
1.4
Urban and Intercity Travel Mode Choice Modeling
1.4.1
Urban Travel Mode Choice Modeling
1.4.2
Intercity Mode Choice Models
1.5
Description of the Course
1.6
Organization of the Course Structure
2
Elements of the Choice Decision Process
2.1
Introduction
2.2
The Decision Maker
2.3
The Alternatives
2.4
Attributes of Alternatives
2.5
The Decision Rule
3
Utility-based Choice Theory
3.1
Basic Construct of Utility Theory
3.2
Deterministic Choice Concepts
3.3
Probabilistic Choice Theory
3.4
Components of the Deterministic Portion of the Utility Function
3.4.1
Utility Associated with the Attributes of Alternatives
3.4.2
Utility ‘Biases’ Due to Excluded Variables
3.4.3
Utility Related to the Characteristics of the Decision Maker
3.4.4
Utility Defined by Interactions between Alternative Attributes and Decision Maker Characteristics
3.5
Specification of the Additive Error Term
4
The Multinomial Logit Model
4.1
Overview Description and Functional Form
4.1.1
The Sigmoid or S shape of Multinomial Logit Probabilities
4.1.2
The Equivalent Differences Property
4.2
Independence of Irrelevant Alternatives Property
4.2.1
The Red Bus/Blue Bus Paradox
4.3
Example: Prediction with Multinomial Logit Model
4.4
Measures of Response to Changes in Attributes of Alternatives
4.4.1
Derivatives of Choice Probabilities
4.4.2
Elasticities of Choice Probabilities
4.5
Measures of Responses to Changes in Decision Maker Characteristics
4.5.1
Derivatives of Choice Probabilities
4.5.2
Elasticities of Choice Probabilities
4.6
Model Estimation: Concept and Method
4.6.1
Graphical Representation of Model Estimation
4.6.2
Maximum Likelihood Estimation Theory
4.6.3
Example of Maximum Likelihood Estimation
5
Data Assembly and Estimation of Simple Multinomial Logit Models
5.1
Introduction
5.2
Data Requirements Overview
5.3
Sources and Methods for Traveler and Trip Related Data Collection
5.3.1
Travel Survey Types
5.3.2
Sampling Design Considerations
5.4
Methods for Collecting Mode Related Data
5.5
Data Structure for Estimation
5.6
Application Data for Work Mode Choice in the San Francisco Bay Area
5.7
Estimation of MNL Model with Basic Specification
5.7.1
Informal Tests
5.7.2
Overall Goodness-of-Fit Measures
5.7.3
Statistical Tests
5.8
Value of Time
5.8.1
Value of Time for Linear Utility Function
5.8.2
Value of Time when Cost is Interacted with another Variable
5.8.3
Value of Time for Time or Cost Transformation
6
Model Specification Refinement: San Francisco Bay Area Work Mode Choice
6.1
Introduction
6.2
Alternative Specifications
6.2.1
Refinement of Specification for Alternative Specific Income Effects
6.2.2
Different Specifications of Travel Time
6.2.3
Including Additinal Decision Maker Related Variables
6.2.4
Including Trip Context variables
6.2.5
Interactions between Trip maker and/or Context Characteristics and Mode Attributes
6.2.6
Additional Model Refinement
6.3
Market Segmentation
6.3.1
Market Segmentation Tests
6.3.2
Market Segmentation Example
6.4
Summary
7
San Francisco Bay Area Shop / Other Mode Choice
7.1
Introduction
7.2
Specification for Shop/Other Mode Choice Model
7.3
Initial Model Specification
7.4
Exploring Alternative Specifications
8
Nested Logit Model
8.1
Motivation
8.2
Formulation of Nested Logit Model
8.2.1
Interpretation of the Logsum Parameter
8.2.2
Disaggregate Direct and Cross-Elasticities
8.3
Nesting Structures
8.4
Statistical Testing of Nested Logit Structures
9
Selecting a Preferred Nesting Structure
9.1
Introduction
9.2
Nested Models for Work Trips
9.3
Nested Models for Shop / Other Trips
9.4
Practical Issues and Implications
10
Multiple Maxima in Nested Logit Estimation (#nesting-optima-chapter)
11
Aggregate Forecasting Assessment, and Application
11.1
Background
11.2
Aggregate Forecasting
11.3
Aggregate Assessment of Travel Mode Choice Models
12
Recent Advances in Discrete Choice Modeling
12.1
The GEV Class of Models
12.2
The MMNL Class of Models
12.3
The Mixed GEV Class of Models
12.4
Summary
References
Published with bookdown
A Self-Instructing Course in Mode Choice Modeling
Chapter 10
Multiple Maxima in Nested Logit Estimation (#nesting-optima-chapter)