Accelerating sustainable mobility

: empirical insights into machine learning-based electric vehicle price prediction

Abstract

This thesis endeavours to discern the primary variables influencing pricing differentials among electric vehicle manufacturers. Leveraging advanced machine learning algorithms, the study scrutinizes the impact of fifteen distinct features that potentially contribute to pricing variations. The models are meticulously trained and constructed on a comprehensive dataset, subsequently tested on an independent sample to ascertain optimal precision and accuracy metrics. While too many studies have successfully implemented this innovative methodology within the domain of internal combustion vehicles, its application to the electric vehicle domain remains a nascent area of inquiry. This pioneering approach, coupled with the evolving landscape of machine learning, holds the promise of delivering dual benefits: affording companies the ability to establish an appropriate pricing spectrum for their vehicles, and providing consumers with access to value-added electric vehicles. The study incorporates a diverse set of regression techniques, including Multiple Linear Regression, Support Vector Machine, Random Forest Regression, Decision Trees and XGBoost Regression. The target variable under consideration is price, characterized by its continuous nature. Consequently, the utilization and implementation of regression methodologies exclusively aligns with the nature of the output variable. In the domain of electric vehicles (EVs), research focused on employing machine learning for pricing determination has been relatively limited in its momentum and significance within the automotive industry. Despite the anticipated proliferation of these battery-powered vehicles driven by global imperatives for carbon neutrality, their widespread adoption is still in its early stages in the 21st century.

Date of Award	24 Jun 2024
Original language	English
Awarding Institution	Universidade Católica Portuguesa
Supervisor	Pedro Afonso Fernandes (Supervisor)