Exploring the Effects of Sampling Locations for Calibrating the Huff Model Using Mobile Phone Location Data

Shiwei Lu, Shih-Lung Shaw, Zhixiang Fang, Xirui Zhang and Ling Yin
Additional contact information
Shiwei Lu: State Key Laboratory of Information Engineering in Surveying, Mapping and Remote Sensing, Wuhan University, 129 Luoyu Road, Wuhan 430079, China
Shih-Lung Shaw: State Key Laboratory of Information Engineering in Surveying, Mapping and Remote Sensing, Wuhan University, 129 Luoyu Road, Wuhan 430079, China
Zhixiang Fang: State Key Laboratory of Information Engineering in Surveying, Mapping and Remote Sensing, Wuhan University, 129 Luoyu Road, Wuhan 430079, China
Xirui Zhang: Information Center of Urban Planning, Land & Real Estate of Shenzhen Municipality, 8007 Hongli West Road, Shenzhen 518040, China
Ling Yin: Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, 1068 Xueyuan Road, Shenzhen 518005, China

Sustainability, 2017, vol. 9, issue 1, 1-18

Abstract: The introduction of the Huff model is of critical significance in many fields, including urban transport, optimal location planning, economics and business analysis. Moreover, parameters calibration is a crucial procedure before using the model. Previous studies have paid much attention to calibrating the spatial interaction model for human mobility research. However, are whole sampling locations always the better solution for model calibration? We use active tracking data of over 16 million cell phones in Shenzhen, a metropolitan city in China, to evaluate the calibration accuracy of Huff model. Specifically, we choose five business areas in this city as destinations and then randomly select a fixed number of cell phone towers to calibrate the parameters in this spatial interaction model. We vary the selected number of cell phone towers by multipliers of 30 until we reach the total number of towers with flows to the five destinations. We apply the least square methods for model calibration. The distribution of the final sum of squared error between the observed flows and the estimated flows indicates that whole sampling locations are not always better for the outcomes of this spatial interaction model. Instead, fewer sampling locations with higher volume of trips could improve the calibration results. Finally, we discuss implications of this finding and suggest an approach to address the high-accuracy model calibration solution.

Keywords: big data; mobile phone location data; spatial interaction model; human dynamics (search for similar items in EconPapers)
JEL-codes: O13 Q Q0 Q2 Q3 Q5 Q56 (search for similar items in EconPapers)
Date: 2017
References: View references in EconPapers View complete reference list from CitEc
Citations: View citations in EconPapers (1)

Downloads: (external link)
https://www.mdpi.com/2071-1050/9/1/159/pdf (application/pdf)
https://www.mdpi.com/2071-1050/9/1/159/ (text/html)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:gam:jsusta:v:9:y:2017:i:1:p:159-:d:88470

Access Statistics for this article

Sustainability is currently edited by Ms. Alexandra Wu

More articles in Sustainability from MDPI
Bibliographic data for series maintained by MDPI Indexing Manager ().