Predicting Used Cars Prices in the Egyptian Market: A Machine Learning Approach

Valuing used cars in the Egyptian market poses a significant challenge due to the scarcity of new vehicles and high inflation. Consequently, used cars dominated the Automotive market, causing price fluctuations and creating a complex pricing environment. Automating the valuation process presents a valuable opportunity to save time and minimize pricing errors. This study explores the use of machine learning models to predict used car prices using data scrapped from popular online listings websites. It compares the performance of top-performing algorithms, according to previous research, including XGBoost, Random Forest Regression, and Bagging Regressor when trained on price-range segmented data versus the full dataset to determine the optimal model. Results indicated that applying price-range segmentation reduced the Mean Absolute Error (MAE) by 23% compared to training without segmentation. XGBoost outperformed other models, achieving an MAE of 47,478, compared to 48,829 for both Random Forest Regression and Bagging Regressor.