An Investigation of the Impact of Amenities on Condominium Apartment Prices in Colombo, Sri Lanka with Machine Learning

Abstract

Condominium prices of a location are heavily impacted by the availability of neighbourhood amenities, which attract people to that location. Studies that employ machine learning to evaluate the impacts of amenities on condominium prices are a few, despite indications of increased accuracy. Hedonic models are used in the majority of current studies on housing pricing. Yet, the application of machine learning will improve accuracy and be more effective for the identification of multi-collinearity and non-linear relationships between property prices and the availability of amenities for a wide range of parameters. The aim of this study is to investigate the possible relationship between urban amenities surrounding condominium apartments and their market prices. The study uses data from the Google Maps Platform to examine the link between neighbourhood amenities and the prices of 236 condominiums in Colombo, Sri Lanka. 56 significant amenity characteristics were found using the eXtreme gradient boosting (XGB) algorithm, and the variety of correlations between amenities and condominium prices as restricted positive, accelerated positive, crooked, humped, and negative was explained. Results demonstrated that while a beautiful urban environment requires a variety of facilities, the popularity and other attributes of amenities affect condominium prices in numerous non-linear ways. Therefore, public and private organizations must work together to create integrative strategies that enhance and maintain the variety and accessibility of urban amenities.