MACHINE LEARNING FOR FUEL TYPE CLASSIFICATION: INSIGHTS FROM THE 2023 TELANGANA VEHICLE SALES DATA SET

Abstract

The growing emphasis on energy efficiency and environmental sustainability has increased the importance of accurate fuel type classification within the transportation sector. Conventional fuel classification methods typically rely on rule-based systems or simple algorithms that use limited vehicle attributes, such as engine size or weight. While these approaches are easy to implement, they often fail to capture the complexity of real-world vehicular data and lack adaptability to emerging vehicle technologies and alternative fuel types. Moreover, manually defined rules can introduce bias and overlook hidden patterns, leading to reduced classification accuracy in dynamic automotive environments. To address these challenges, this study proposes a machine learning–based fuel type classification system that leverages the Telangana Vehicle Sales 2023 dataset. The proposed approach utilizes advanced learning techniques, including deep neural networks and ensemble-based models, to automatically identify complex relationships among diverse features. The system considers a rich set of attributes such as engine specifications, vehicle weight, emission characteristics, geographical location, and socio-economic factors. By integrating these multidimensional features, the model delivers a more precise and robust classification of vehicle fuel types. The proposed system enhances understanding of regional vehicle usage patterns and supports data-driven decision-making for policymakers and stakeholders. Ultimately, this research contributes to sustainable transportation planning by enabling more accurate fuel classification and promoting environmentally responsible mobility solutions.