The effect of adopting micro and macro-economic variables on real estate price prediction models using ANN: A systematic literature review

Abstract

The trend in real estate price estimation tends towards the adoption of artificial intelligence (AI) where micro variables related to real estates have been widely adopted. Whereas, macro-economic variables also have a significant role in price determination. This study, therefore, examined the trends in both micro and macro-economic variable adoption in Artificial Neural Network (ANN) related researches within the past two decades with a view to assessing their impact on the models performance. This is intended to expose the gap in the literature, in order to guide future researches in the field of AI application in price prediction. Using R2 in error measurement as a basis, the study revealed that researches that adopted macro-economic variables had 100% of the R2 values above 0.95, while studies that adopted micro variables recorded only 23% above 0.9, and 54% of their R2 to be between 0.8 to 0.9. Nevertheless, studies that adopted both variables stood at the average with 50% of their R2 readings above 0.9 and 33% was between 0.8 and 0.9. Thus, the study concludes that there is the need for future AI-related studies to explore a combination of both variables in order to avoid the two extremes.