Publication Type

Conference Proceeding Article

Version

publishedVersion

Publication Date

6-2022

Abstract

In this research, a taxi travel time based Geographically Weighted Regression model (GWR) is proposed and utilized to model the public housing price in the case study of Singapore. In addition, a comparison between the proposed taxi data driven GWR and other models, such as ordinary least squares model (OLS), GWR model based on Euclidean distance and GWR model based on public transport travel time, have also been carried out. Results indicates that taxi travel time based GWR model has better fitting performance than the OLS model, and slightly better than the Euclidean distance-based GWR model, however, it is not as good as the GWR model based on public transport travel time according to the metric of Adjusted R2. These experiments indicate that the public transport travel time may has a major part to play in modeling the public housing resale prices compared to taxi travel time or driving time, and both the taxi travel time and public transport travel time can better explain the public housing resale prices in Singapore compared to Euclidean distance in the GWR modeling.

Keywords

Hedonic model, GWR, Public housing prices, Taxi travel time

Discipline

Computer Engineering

Research Areas

Intelligent Systems and Optimization

Publication

Proceedings of the 29th International Conference on Geoinformatics, Beijing, China, 2022 June 15 - 18

Identifier

10.1109/Geoinformatics57846.2022.9963833

Publisher

IEEE

City or Country

Beijing, China

Additional URL

https://doi.org/10.1109/Geoinformatics57846.2022.9963833

Share

COinS