Machine learning approaches for electricity price prediction: A case study on the day - ahead market

Department of Power Systems and Management, Technical University of Cluj-Napoca, Cluj-Napoca, Romania

^* Corresponding author: This email address is being protected from spambots. You need JavaScript enabled to view it.

Published online: 11 June 2026

Abstract

This paper presents a comparative study on the use of machine learning models for predicting electricity prices on the Day-Ahead Market (DAM) in Romania, using the Orange Data Mining platform. To validate the study, real hourly data for June 2025 was collected and processed. Based on this data, three different machine learning algorithms were trained and evaluated: linear regression, Random Forest, and Neural Networks. The experimental workflow encompassed data correlation and preprocessing, the selection of salient features, and the evaluation of models based on key metrics such as root mean square error and the coefficient of determination R². The findings indicate that the Random Forest model exhibits a marked superiority over both linear regression and neural networks, achieving an R² coefficient of 0.791 and a minimal prediction error. The study underscores the significance of employing flexible models that are equipped to capture complex and non-linear relationships that are characteristic of the energy market. Additionally, it emphasizes the essential role of crucial features in enhancing the accuracy of predictions. The Orange platform has demonstrated its efficacy and accessibility in facilitating the expeditious development and comparison of predictive models. The future direction of this research includes the expansion of the dataset, the integration of additional relevant variables, and the exploration of advanced neural network architectures.