High-Accuracy Car Price Prediction and Analysis of Influential Factors Using a Comparative Approach of Ensemble Machine Learning Algorithms

Accurate car price prediction is of high importance as a key issue in economic markets and for consumer decision-making. The objective of this research is to present a comprehensive comparative study on the performance of modern ensemble machine learning algorithms in predicting car prices, as well as to identify and analyze the key factors influencing it. In this study, a public dataset containing technical specifications and appearance characteristics of vehicles was utilized. After a preprocessing stage, which included data cleaning and converting categorical variables into a numerical format, the data was prepared for modeling. Three powerful algorithms—Random Forest, Extreme Gradient Boosting (XGBoost), and Light Gradient Boosting Machine (LightGBM)—were trained and evaluated for this purpose. The evaluation results, based on the Coefficient of Determination (𝑅 ۲) metric, indicated that the Random Forest model achieved the highest performance among the models studied, with a score of ۰.۹۶. This result demonstrates a significant improvement over previous studies, which had at best achieved a coefficient of determination of approximately ۰.۴۰. Furthermore, the feature importance analysis revealed that variables such as curb weight, engine displacement, and engine power (horsepower) have the most significant impact on determining the final price. The findings of this research not only prove the high efficacy of ensemble methods for this problem but also provide a valuable guide for future research by presenting a high-accuracy benchmark and an analysis of influential factors.

کلیدواژه ها:

Car Price Prediction ، Machine Learning ، Ensemble Algorithms ، Random Forest ، Feature Importance Analysis