New technologies such as smart phones and web applications constantly collect data on individuals' trip-making and travel patterns. Efforts at using these "Big data" products, to date, have focused on using them to expand or inform traditional travel demand modeling frameworks; however, it is worth considering if a new framework built to maximize the strengths of big data would be more useful to policy makers and planners.
In this presentation Greg Macfarlane will present a discussion on elements of travel models that could quickly benefit from big data and concurrent machine learning techniques, and results from a preliminary application of a prototype framework in Asheville, North Carolina.
Dr. Macfarlane is an analyst in the Systems Analysis Group of WSP | Parsons Brinckerhoff, developing and applying advanced travel demand models. His research and expertise includes trip-based models, activity-based models, integrated land-use/transport models, and micro-simulation of both travel demand and supply. He is an advocate of using open and robust software design principles in public projects, and is developing a pattern-based demand model as an associate with Transport Foundry. Greg is an alumnus of BYU and received graduate degrees from Georgia Tech.