One of transportation research topic is detecting trip purpose. Given a collection of GPS mobility records, researchers endeavored to infer useful information such as trip, travel mode, and trip purpose. Obtaining these attributes will help researcher in transportation modelling. Â This work proposed an approach in defining a trip or a trip segmentation which is a part of trip purpose problem as well as inferring the trip purpose. By Utilizing Dbscan clustering algorithm, decision tree, and some useful features, we are able to detect the trips and their purposes as well as building the model to automate the trip derivation.
Copyrights © 2018