This project is in collaboration with the National Center for Smart Growth and focuses on collecting new sources of data and exploring machine learning techniques for predicting ridership across metro stations in the D.C. metropolitan area. The new data and methods are being used to replace a previous origin-destination land use ridership model (OD LURM), which is essentially a type of spatial interaction model deployed in the context of transit planning. The work also produces a reproducible pipeline for cleaning and processing dozens of variables for use within models.

Supported by the Washington Metropolitan Area Transit Authority (WMATA)