Thursday, March 11, 2021
FDOT Central Office Announces Modeling Training Courses
The FDOT Forecasting and Trends Office will offer two online modeling training webinar series in upcoming weeks. To sign up, please email Terry Corkery at terrence.corkery@dot.state.fl.us.
- Executive Summary Modeling Webinar - March 29 and 30, 2021
Monday, March 29, 2021, 1:00 PM to 4:30 PM Eastern, and Tuesday, March 30, 2021, 8:30 AM to noon Eastern
This seven-hour beginning-level course is designed for managers and other non-modelers who review model outputs. The workshop provides an overview of the transportation planning process, travel demand forecasting methodologies, and FSUTMS modules. Participants learn the underlying theories of modeling and see the benefits of the models' abilities to produce presentation graphics for reports and meetings. (see workshop agenda)
- Advanced Python Programming: Choice Modeling - March 29 to April 9, 2021
Lunch and learn: Noon to 1:30 PM every weekday for two weeks
This ten-part webinar series provides a basic introduction on the use of Python for the estimation and application of discrete choice models. It includes a brief review of the fundamentals of discrete choice theory, but focuses mostly on the practical tools for working with these models in Python, including Numpy, Pandas, and Larch. It includes modules on data preparation, non-linear optimization, and model estimation, analysis, and application for multinomial logit and nested logit models. The course assumes that students are already somewhat familiar with the methodologies and mathematical tools of transportation modeling, and focuses on the how these models are constructed and used within Python. Sections of the course offer a concrete, practical parallel to the freely available Self Instructing Course in Mode Choice Modeling: Multinomial and Nested Logit Models, which is recommended as companion reading for participants who are less familiar with discrete choice theory.
Pre-requisites:
- Basic familiarity with Python, at least to the level covered in the “Python for Transportation Modeling” course
- General familiarity with the theory of discrete choice models is suggested but not required
Agenda
- Discrete Choice Theory Review
- Estimation of Choice Model Parameters
- Statistical Evaluation of Models
- Managing Data for Choice Models
- Mode Choice Model Development
- Nested Logit Models
- Developing and Using Logsums
- Destination Choice Models
- Moving from Estimation to Application
