Electric Vehicle Price Prediction Using Real-World Online Listings: An Explainable Machine Learning Approach

Abstract

The recent gains in the adoption of electric vehicles (EVs) have triggered the emergence of secondary markets concerning EVs, where the determination of prices accurately and transparently is an extreme challenge. In the current study, we propose an explainable machine learning model to predict the advertised price (in Jordanian dinars) of used EVs, using actual online advertisements obtained in the Jordanian market. An EV-oriented and custom-made dataset of 2,165 listings was used that specifically includes battery capacity and driving range along with the typical vehicle features. The coefficient of determination (R²), mean absolute error, and root mean squared error were used to assess the performance of three tree-based regressors: CatBoost Regressor, Extreme Gradient Boosting Regressor, and Random Forest Regressor trained with a leakage-safe pipeline and evaluated using grouped, repeated cross-validation. SHapley Additive exPlanations were applied to the fitted models to quantify the global contribution of each feature to price predictions and to assess the stability of these contributions across validation folds, providing a transparent account of the price-determination mechanism. The empirical data indicate that battery capacity and driving range are the most dominant EV price determinants that continue to overshadow conventional predictors of price, including brand, mileage, and year. The soundness of such findings is supported by cross-model explainability analyses. As a result, the proposed framework enhances pricing transparency and enables informed decision-making in young EV markets.



Received: 13 March 2026 | Revised: 17 August 2026 | Accepted: 21 September 2026



Conflicts of Interest

The authors declare that they have no conflicts of interest to this work.



Data Availability Statement

The data that support the findings of this study are openly available in Mendeley Data at https://doi.org/10.17632/ddcz486x5t.2.



Author Contribution Statement

Nisrean Jaber Thalji: Conceptualization, Methodology, Formal analysis, Investigation, Data curation, Writing – original draft, Writing – review & editing, Visualization, Supervision, Project administration. Aryaf Al-Adwan: Methodology, Validation, Investigation, Writing – review & editing. Tariq Bishtawi: Software, Validation, Writing – review & editing. Hassan Kassem: Software, Validation, Writing – review & editing.