International Choice Modelling Conference, International Choice Modelling Conference 2017

Font Size: 
Evaluating a data-driven approach for choice set identification
Danique Ton, Dorine Duives, Oded Cats, Serge Hoogendoorn

Last modified: 28 March 2017


Given the introduction of new data collection methods, like GPS, which provide detailed spatial and temporal data on the movements of individuals, it is now possible to derive commonly chosen choice sets in a data-driven manner. To understand the potential of the data-driven path identification (DDPI) approach proposed in this paper, the method is compared to the Breadth-first search link elimination approach and k-shortest paths approach in a bicycle route chocie case study in the inner-city of Amsterdam. The comparison is made both on the resulting choice sets and model estimates using the Path Size Logit model.




Conference registration is required in order to view papers.