Predicting Real Estate Prices Using Machine Learning in Abu Dhabi

Abstract

     Traditionally, real estate prices were determined based on demand and supply. As the real estate market was  unregulated and underdeveloped, brokers and real estate builders had an upper hand in determining the unit prices of residential houses in Abu Dhabi. A pricing gap was eventually noticed. This was a challenge. There is a delay in updating the real estate websites and portal information. Therefore, the need for accurate forecasting of prices has become urgent. With a variety of use case scenarios for machine learning concepts, this paper is dedicated to using the concepts of machine learning to predict the real estate prices of Aldar in the Abu Dhabi region, which comprises 511 residential units, 15 retail shops, and one community center. Decision tree, random forest, support vector machines, and K-nearest neighbors (KNN) algorithms were used to identify which one is better for forecasting these real estate prices. Comparing the generated models, the random forest is the best-performing model, followed by support vector regression, and the decision tree model is the least-performing model.