TY - CHAP
T1 - The superstation representation of metro networks for overcoming data availability issues of station-to-station origin destination pairs – An application on the London underground
AU - Nádudvari, Tamás
AU - Liu, Ronghui
AU - Balijepalli, Chandra
AU - Fu, Qian
PY - 2019/12
Y1 - 2019/12
N2 - Information on metro passengers’ route choices is important to public transport operators. Recently, smart cards have been introduced in many metropolises, which, however, do not reveal explicitly the actual routes of the cardholders. Research literature shows that a finite mixture model (FMM) may provide an efficient approach to inferring the route choices; still, it can be rendered ineffective when the data sample size is small. To overcome this issue, we introduce the concept of ‘superstations’, which is a group of stations from/to which passengers have similar route choice patterns. From that, a larger data sample can be available for a superstation-to-superstation origin destination pair, so as to regain a functional FMM and enhance its fitness for inferring the passengers’ route choices. We test the proposed methodology on the London Underground. Results show that the FMM applied on the superstation representation presents a better performance than using a simple station-to-station representation.
AB - Information on metro passengers’ route choices is important to public transport operators. Recently, smart cards have been introduced in many metropolises, which, however, do not reveal explicitly the actual routes of the cardholders. Research literature shows that a finite mixture model (FMM) may provide an efficient approach to inferring the route choices; still, it can be rendered ineffective when the data sample size is small. To overcome this issue, we introduce the concept of ‘superstations’, which is a group of stations from/to which passengers have similar route choice patterns. From that, a larger data sample can be available for a superstation-to-superstation origin destination pair, so as to regain a functional FMM and enhance its fitness for inferring the passengers’ route choices. We test the proposed methodology on the London Underground. Results show that the FMM applied on the superstation representation presents a better performance than using a simple station-to-station representation.
KW - Data availability
KW - Finite mixture model
KW - Route choice
KW - Smart card
KW - Superstation
UR - http://www.scopus.com/inward/record.url?scp=85083889265&partnerID=8YFLogxK
UR - https://scholars.cityu.edu.hk/en/publications/publication(e130b841-fc2d-4d56-a2ff-e59c906a1842).html
M3 - Chapter
AN - SCOPUS:85083889265
SN - 9789881581488
T3 - Proceedings of the 24th International Conference of Hong Kong Society for Transportation Studies, HKSTS 2019: Transport and Smart Cities
SP - 287
EP - 294
BT - Proceedings of the 24th International Conference of Hong Kong Society for Transportation Studies, HKSTS 2019
A2 - Chow, Andy H.F.
A2 - Lo, S.M.
A2 - Li, Lishuai
PB - Hong Kong Society for Transportation Studies
T2 - 24th International Conference of Hong Kong Society for Transportation Studies: Transport and Smart Cities, HKSTS 2019
Y2 - 14 December 2019 through 16 December 2019
ER -