Gradient boosting regression approach for housing unit price prediction

o purchase a house is one of the biggest financial goals for everyone. However, accurate and prompt housing unit price (HUP) prediction is crucial for both the real estate industry and investors. This study proposes a HUP prediction model based on gradient boosting regression (GBR). The proposed GBR model was compared with the following investigating methods: adaptive boosting (AdaBoost), k-nearest neighbour (KNN), decision tree (DT), random forest (RF), and support vector machine (SVM). The proposed GBR method demonstrated superior predictive performance over five state-of-the-art methods (AbaBoost, KNN, DT, RF, and SVM) when evaluated using a real dataset. This was obvious from the mean absolute percentage error (MAPE), coefficient of determination (R2), correlation coefficient (R), and coefficient of variance root mean square error (CVRMSE) employed as model assessment metrics. The results revealed that the GBR had the lowest MAPE (0.017%), CVRMSE (1.968%), and highest R2 (0.993) and R (0.99649) values as compared with the other investigated methods. This confirms the proposed GBR method’s strength for reliable and efficient HUP prediction.