Around the world, traffic congestion is a major issue that affects everyday commutes, economic activity, and the sustainability of the environment. The goal of this research project is to use machine learning approaches to address the problem of traffic congestion. Through the utilization of an extensive dataset that encompasses several spatiotemporal parameters, including date, time, weather, and holiday indicators, the research creates predictive models that enable precise forecasting of traffic congestion levels. In order to determine how well machine learning predicts congestion dynamics, the research uses multiple regression methods, such as MLP Regressor, Stacking Regressor, and SVR. Historical traffic volume data is used to train and assess the models, making it possible to identify the main variables impacting patterns of congestion. The findings demonstrate how weather, time of day, and holidays all have a major impact on traffic congestion. Transportation authorities and urban planners may enhance overall urban mobility, optimise infrastructure utilisation, and regulate traffic flows proactively by leveraging the predictive capabilities of machine learning models. This work advances intelligent transport systems by offering a strong framework for anticipating and controlling traffic congestion. The results highlight the significance of utilising data-driven methodologies to tackle intricate urban issues, ultimately cultivating more effective and sustainable urban settings.