Towards Smarter Climate Monitoring: CO2 Emission Predictions through Machine Learning

The development of effective prediction models for emissions monitoring and control is essential due to the growing amounts of CO₂ emissions from automobiles, which greatly contribute to environmental pollution as well as global warming. Conventional techniques for calculating CO2 emissions frequently depend on empirical calculations and static presumptions, which might not adequately represent the intricate connections between emission rates and vehicle attributes. This paper presents a machine learning-based method to forecast CO₂ emissions based on important vehicle metrics including engine size (L), cylinder count and fuel use (L/100 km) in order to solve Environmental problem. In order to refine the dataset before training several models, such as Linear Regression, Random Forest, Support Vector Regression (SVR) and K-Nearest Neighbors (KNN), the study uses exploratory data analysis (EDA) and data preprocessing. The Random Forest model was the best option for real-time predictions, with the greatest degree of precision (0.9609), according to performance assessment using the R2 score. In order to provide customers with an easy-to-use and efficient tool for making decisions in the automotive and environmental policy sectors, a Streamlit-based user interface (UI) was created that enables users to enter car specs and instantaneously get CO₂ emission forecasts. This paper demonstrates how machine learning may improve pollution control methods encourage environmentally friendly transportation and support legal compliance.