Determinants of housing rental prices in Seoul: applying explainable AI

Tae-Young Kim, Eunil Park and Doojin Ryu

Spatial Economic Analysis, 2025, vol. 20, issue 2, 312-332

Abstract: Though previous analyses of housing rental prices rely on hedonic pricing or ordinary least squares models, they cannot fully account for the nonlinearity and complexity inherent in the determination of market prices. We exploit machine-learning techniques to accurately understand the determinants of housing rental prices and emphasise explainable artificial intelligence (XAI) to overcome the limitations of classical machine learning with black-box features. To analyse the determinants of housing rental prices in Seoul, we apply extreme gradient boosting (XGBoost) and Tree Shapley additive explanations (TreeSHAP) models which learn and explain nonlinear features of the rental prices. XGBoost predicts rental prices more accurately than regressions and is robust to multicollinearity. TreeSHAP identifies local and global features affecting the determination of rental prices. Our analyses identify the individual impacts of various determinant variables on rental prices and underscore the multifaceted influences that these variables exert on housing affordability.

Date: 2025
References: Add references at CitEc
Citations:

Downloads: (external link)
http://hdl.handle.net/10.1080/17421772.2024.2418906 (text/html)
Access to full text is restricted to subscribers.

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:taf:specan:v:20:y:2025:i:2:p:312-332

Ordering information: This journal article can be ordered from
http://www.tandfonline.com/pricing/journal/RSEA20

DOI: 10.1080/17421772.2024.2418906

Access Statistics for this article

Spatial Economic Analysis is currently edited by Bernie Fingleton and Danilo Igliori

More articles in Spatial Economic Analysis from Taylor & Francis Journals
Bibliographic data for series maintained by Chris Longhurst ().