Publication Type

Conference Proceeding Article

Version

acceptedVersion

Publication Date

12-2018

Abstract

The mass rapid transit (MRT) network is playing an increasingly important role in Singapore's transit network, thanks to its advantages of higher capacity and faster speed. Unfortunately, due to aging infrastructure, increasing demand, and other reasons like adverse weather condition, commuters in Singapore recently have been facing increasing unexpected train delays (UTDs), which has become a source of frustration for both commuters and operators. Most, if not all, existing works on delay management do not consider commuters' behavior. We dedicate this paper to the study of commuters' behavior during UTDs. We adopt a data-driven approach to analyzing the six-month' real data collected by automated fare collection system in Singapore and build a classification model to predict whether commuters switch from MRT to other transportation modes because of UTDs.

Keywords

Mass Rapid Transit, unexpected train delays, smart card data, trip chains, individual travel patterns, clustering, DBSCAN, feature engineering, response modeling, feature insights

Discipline

Databases and Information Systems | Numerical Analysis and Scientific Computing | Transportation

Research Areas

Data Science and Engineering

Publication

2018 IEEE International Conference on Big Data: Seattle, WA, December 10-13: Proceedings

First Page

831

Last Page

840

ISBN

9781538650356

Identifier

10.1109/BigData.2018.8622233

Publisher

IEEE Computer Society

City or Country

Los Alamitos, CA

Copyright Owner and License

Authors

Additional URL

https://doi.org/10.1109/BigData.2018.8622233

Download

Share

COinS