Fuel Prediction Model for Driving Patterns Using Machine Learning Techniques

Abstract

In recent years, the demand for fuel efficiency has become a crucial aspect of the automotive industry. Predicting fuel economy accurately is essential for optimizing vehicle performance and reducing environmental impact. This proposed model presents a machine learning-based approach for developing a fuel prediction model using the real dataset. The model aims to predict fuel efficiency for different drivers by considering input features related to driving conditions and driver behavior. The study explores the application of linear regression and support vector regression, to achieve accurate and reliable predictions. The dataset is pre-processed to handle missing values, normalize numerical features, and encode categorical variables. Feature engineering techniques are employed to select the most relevant features and enhance the model's performance. A thorough assessment is carried out utilizing diverse performance measures, including mean squared error and R-squared score, to evaluate the forecasting aptitude of the created models. The linear regression model exhibits exceptional performance, as evidenced by its high R-squared values (0.9963%, approaching 1) and low values for MAE, MSE, and RMSE. The outcomes illustrate the efficacy of the suggested method in precisely forecasting fuel economy for varying drivers for even real-time values. The findings provide valuable insights for vehicle manufacturers, policymakers, and individuals interested in optimizing fuel consumption and reducing greenhouse gas emissions. Overall, this study contributes to the growing body of knowledge in machine learning techniques and reinforces the significance of machine learning techniques in addressing fuel economy challenges with driver’s behaviors.