TRAVEL MODE DETECTION IN NEW YORK CITY USING SMARTPHONES AND GEOGRAPHIC INFORMATION SYSTEMS (GIS)
Abstract
A custom classification algorithm that accounts for the nuances of travel (e.g., underground movements, frequent mode changes, and severe urban canyon effects) in New York City combines a probabilistic random forest classifier with a deterministic GIS-based post-processing procedure to address the challenges of travel mode detection in the city.
This paper has been withdrawn.