Geographical Huff Model Calibration using Taxi Trajectory Data

Shuhui Gong, John Cartlidge, Yang Yue, Guoping Qiu, Qingquan Li, Jingyu Xin

Research output: Chapter in Book/Report/Conference proceedingConference Contribution (Conference Proceeding)

6 Citations (Scopus)
268 Downloads (Pure)

Abstract

The widely used Huff model is designed to estimate the spatial probability distribution of shopping centre patronage based on a shopping centre’s attractiveness and the cost of a customer’s travel. Here, we calibrate the Huff model for the city of Shenzhen, China, using GPS taxi trajectory data for one million taxi journeys. Using Geographical Weighted Regression to fit the model, we show that there is significant geographical variation in best estimates of the Huff parameters of attractiveness and cost. To explain this variation, we use open-source house price sales’ data as a proxy for customers’ wealth in each region. Regression results demonstrate a significant linear relationship between localised house prices and the Huff model parameter of attractiveness, suggesting that wealthy customers are more sensitive to shopping centre attractiveness than customers with less wealth. We present this as a novel discovery.
Original languageEnglish
Title of host publicationProceedings of 10th ACM SIGSPATIAL International Workshop on Computational Transportation Science
Subtitle of host publicationRedondo Beach, CA, USA, November 7–10, 2017 (IWCTS’17)
PublisherAssociation for Computing Machinery (ACM)
Pages30-35
Number of pages6
ISBN (Electronic)9781450354912
DOIs
Publication statusPublished - 10 Nov 2017

Keywords

  • Shopping behavior
  • Taxi data
  • House price data
  • GWR
  • OLS

Fingerprint

Dive into the research topics of 'Geographical Huff Model Calibration using Taxi Trajectory Data'. Together they form a unique fingerprint.

Cite this