Does artificial intelligence enhance house price forecasting accuracy? – a literature review

• Main Authors: Sa’at, Nurul Fazira, Adi Maimun, Nurul Hana
• Format: Article
• Language: English
• Published: Universiti Teknologi Malaysia 2019
• Subjects: TH434-437 Quantity surveying
• Online Access: http://eprints.utm.my/id/eprint/87273/1/NurulHanaAdiMaimun2019_DoesArtificialIntelligenceEnhanceHouse.pdf; http://eprints.utm.my/id/eprint/87273/; https://www.utm.my/intrest/files/2019/12/6_DOES-ARTIFICIAL-INTELLIGENCE-ENHANCE-HOUSE-PRICE-FORECASTING-ACCURACY-%E2%80%93-A-LITERATURE-REVIEW.docx.pdf
• Institution: Universiti Teknologi Malaysia

The fast-changing housing industry demands the adoption of advanced approaches to valuation for a quick, reliable and accurate result. The traditional approach for forecasting house prices, called the Hedonic Pricing Model (HPM), is problematic given its inaccuracy due to problems with heteroskedasticity and multicollinearity among variables in the model. Recently, there is increasing attention in the application of Artificial Intelligence (AI) to forecast house prices. AI, through Artificial Neural Network (ANN), addresses the shortcomings of HPM. Hence, this paper aims to critically review previous studies on the ability of ANN as a substituted model for HPM in forecasting house prices. Various secondary sources were involved due to extracting various documentary data. It was concluded that the application of AI-enhanced forecasting is accurate. This was demonstrated through the superior predictive performance of ANN compared to HPM.