This project applies machine learning algorithms to predict the real estate unit’s values in the emirates of Ajman. Adopting machine learning techniques to predict real estate unit prices helps us produce more accurate valuations. It allows us to use historical data that can’t be used in traditional valuation techniques.
Datasets from the Department of Land & Real Estate were used to run an exploratory analysis to explore the most critical factors influencing real estate unit prices in Ajman. It shows that the unit area is the most significant factor affecting Ajman's real estate unit prices. It also shows how the real estate market is trending and the effect of Covid-19 on the real estate market in Ajman.
Predictive Analysis was applied using the R programming language to predict real estate unit prices. Three models were used: Multiple Linear Regression, Support Vector Regression (SVR), and Gradient Boosting (GB) Model. Stepwise Regression was used to select the most significant attributes that will be used to build the model. Genetic Algorithms were used to tune parameters for the Support Vector Regression and Gradient Boosting Model. The resulting models were then evaluated using R2 and RMSE, where the GB model provided the best results.
This project will help Ajman Land and Real Estate Regulation Department to get more accurate values for real estate unit prices on the spot, which reduces the service time and steps. It also helps create a more reliable real estate market.
Professional Studies (MS)
Department, Program, or Center
Graduate Programs & Research (Dubai)
Al Yazouri, Heba, "Using ML Algorithms to Predict Real Estate Units Valuations in Ajman" (2023). Thesis. Rochester Institute of Technology. Accessed from