Google Maps amenities and condominium prices: Investigating the effects and relationships using machine learning

Abstract

Neighbourhood amenities significantly impact condominium prices and attract population to the area. Despite evidence of improved accuracy, studies that use machine learning to assess the effects of amenities on condominium prices are limited. The purpose of this research is to investigate the relationship between neighbourhood amenities and the prices of 500 condominiums in Bangkok, Thailand using data from Google Maps. An eXtreme gradient boosting (XGB) algorithm identified 36 important amenity factors, while the multiplicity of relationships between amenities and condominium prices as bounded positive, accelerated positive, limited positive, humped and negative was elucidated. Results showed that the popularity and other features of amenities drive condominium prices in several non-linear ways, while an attractive urban environment requires multiple amenities. Public and private organisations in Bangkok should collaborate to develop integrative plans that improve and sustain the diversity and availability of urban amenities.