Courses Taught
Choice Modeling (CE 271)
This course covers the theory, basic and advanced modeling methods, and applications of econometric choice models for analyzing consumer choice behavior; with applications drawn from transportation. Beginning with the basic choice modelling methods such as the multinomial logit and maximum likelihood estimation, the course will delve into advanced model formulations and estimation methods as well as issues of parameter identification in detail. In addition to theories and modelling methods, importance is given to hands-on estimation, specification, and interpretation of choice models on real-life empirical datasets. Issues such as empirical specification, interpretation, causality, behavioural heterogeneity, and endogeneity will be discussed.
Course Content (tentative)
(a) Basics: Individual choice theories, Binary choice models, Unordered multinomial choice models (logit and probit), Maximum likelihood estimation, Goodness of fit, Hypothesis testing, Sampling (choice-based samples, sampling of alternatives), Marginal and elasticity effects, Welfare analysis, Prediction
(b) Multidimensional choice models: Nested logit and Cross-nested logit
(c) Ordered response models: Ordered logit, Ordered probit, Generalized ordered response models
(d) Mixture models: Mixed Logit, Latent class (discrete mixture) models, choice models with cross-sectional and panel data, Simulation-based and EM estimation
(e) Multiple Discrete Choice models: Discrete-continuous models, Kuhn-Tucker demand systems
(f) Unordered probit models: Multinomial probit model, Multivariate probit models, Methods for computing multivariate normal CDF
(g) Models with latent variables: Integrated latent variable and choice models (choice models with psycho-social attitudinal and life-style variables), Models with stochastic variables and random parameters
(h) Parameter identification in basic and advanced choice models: Theoretical and empirical identification
(i) Alternative estimation techniques: Simulated likelihood, Analytic Approximations, Composite Likelihood, EM
Travel Demand Modeling (CE 270)
This is a first graduate course in the theory, development, estimation and application of statistical models for travel demand analysis. Below are the course contents:
Individual travel behavior and aggregate-level travel demand analysis; Alternative approaches to modeling travel demand (aggregate, trip-based approaches and disaggregate, activity-based approaches); Econometric methods for modeling travel demand (development, estimation, and application of statistical models for travel behavior analysis); Linear regression for activity and trip generation (specification, interpretation, estimation, hypothesis testing, market segmentation, non-linear specification, tests on assumptions); Mode choice and destination choice using discrete choice methods (introduction to binary logit and multinomial logit models, contrast with entropy and gravity methods); Traffic assignment/route choice (network equilibrium, system optimum); Model transferability; Microsimulation for activity-based models; Recent advances.
Graduate Courses
Travel Demand Modeling
Discrete Choice Analysis
Sustainable Transportation
Transportation Research Seminar
Traffic Systems Engineering
Undergraduate Courses
Transportation Engineering-1
Transportation Engineering-2
Transportation Laboratory
Professional Development Courses
This course covers the theory, basic and advanced modeling methods, and applications of econometric choice models for analyzing consumer choice behavior; with applications drawn from transportation. Beginning with the basic choice modelling methods such as the multinomial logit and maximum likelihood estimation, the course will delve into advanced model formulations and estimation methods as well as issues of parameter identification in detail. In addition to theories and modelling methods, importance is given to hands-on estimation, specification, and interpretation of choice models on real-life empirical datasets. Issues such as empirical specification, interpretation, causality, behavioural heterogeneity, and endogeneity will be discussed.
Course Content (tentative)
(a) Basics: Individual choice theories, Binary choice models, Unordered multinomial choice models (logit and probit), Maximum likelihood estimation, Goodness of fit, Hypothesis testing, Sampling (choice-based samples, sampling of alternatives), Marginal and elasticity effects, Welfare analysis, Prediction
(b) Multidimensional choice models: Nested logit and Cross-nested logit
(c) Ordered response models: Ordered logit, Ordered probit, Generalized ordered response models
(d) Mixture models: Mixed Logit, Latent class (discrete mixture) models, choice models with cross-sectional and panel data, Simulation-based and EM estimation
(e) Multiple Discrete Choice models: Discrete-continuous models, Kuhn-Tucker demand systems
(f) Unordered probit models: Multinomial probit model, Multivariate probit models, Methods for computing multivariate normal CDF
(g) Models with latent variables: Integrated latent variable and choice models (choice models with psycho-social attitudinal and life-style variables), Models with stochastic variables and random parameters
(h) Parameter identification in basic and advanced choice models: Theoretical and empirical identification
(i) Alternative estimation techniques: Simulated likelihood, Analytic Approximations, Composite Likelihood, EM
Travel Demand Modeling (CE 270)
This is a first graduate course in the theory, development, estimation and application of statistical models for travel demand analysis. Below are the course contents:
Individual travel behavior and aggregate-level travel demand analysis; Alternative approaches to modeling travel demand (aggregate, trip-based approaches and disaggregate, activity-based approaches); Econometric methods for modeling travel demand (development, estimation, and application of statistical models for travel behavior analysis); Linear regression for activity and trip generation (specification, interpretation, estimation, hypothesis testing, market segmentation, non-linear specification, tests on assumptions); Mode choice and destination choice using discrete choice methods (introduction to binary logit and multinomial logit models, contrast with entropy and gravity methods); Traffic assignment/route choice (network equilibrium, system optimum); Model transferability; Microsimulation for activity-based models; Recent advances.
Graduate Courses
Travel Demand Modeling
Discrete Choice Analysis
Sustainable Transportation
Transportation Research Seminar
Traffic Systems Engineering
Undergraduate Courses
Transportation Engineering-1
Transportation Engineering-2
Transportation Laboratory
Professional Development Courses
- A two-day workshop on activity-based travel demand models to consultants and stakeholders of the Florida Department of Transportation (Co-taught with Drs. Siva Srinivasan, John Bowman, and Joel Freedman)
- A three-day short course on "Discrete Choice Theory and Modeling Applications in Transportation" (Co-taught with Dr. Rajesh Paleti)
- A five-day SPARC course on "Advanced Choice Modelling Methods in an Evolving Urban Travel Behaviour Landscape" (Co-taught with Prof. Chandra Bhat and Prof. Karthik Srinivasan)
