Car Prices Using Data Mining Techniques An applied study on a UK cars dataset

Abstract

The study aims to examine the effectiveness of the XGBoost regression algorithm and
compare it with the Random Forest (RF) algorithm, the Support Vector regression (SVR)
algorithm, and the Deep Neural Networks (DNN) algorithm to predict car prices in the United
Kingdom, the study found that the Random Forest (RF) algorithm is appropriate and effective for
accurately estimating car prices, and therefore can be relied upon to improve and rationalize
decisions of car buyers and seller, This conclusion is based on the RF algorithm achieving the
highest coefficient of determination (R²) of 95.90 % and the lowest Root Mean Squared Error
(RMSE) of 1946.07, compared to the XGBoost regression algorithm, the Support Vector regression
(SVR) algorithm, and the Deep Neural Networks (DNN) algorithm.
Keywords: Data mining; XGBoost regression algorithm; Random Forest (RF) algorithm; Support
Vector regression (SVR) algorithm; Deep Neural Networks (DNN) algorithm