Deep Learning-Based Urban Traffic Congestion Forecasting Using Transformer Models

Traffic congestion has become a major issue in urban areas due to rapid population growth and increased vehicle usage. Efficient traffic prediction is essential for improving transportation systems and reducing delays, fuel consumption, and environmental impact. This research focuses on predicting traffic congestion using machine learning techniques based on historical traffic data. The dataset used in this study contains time-based traffic values, which are analyzed to identify patterns and trends. In this project, a Linear Regression model is implemented to establish a relationship between time and traffic values, enabling the prediction of future traffic conditions. The model is trained and tested using a dataset, and the results are visualized through a graph comparing actual and predicted values. The findings indicate that the model can effectively capture general traffic trends, although it may have limitations in handling sudden fluctuations. This study highlights the importance of data-driven approaches in traffic management. Future work can focus on improving prediction accuracy using advanced techniques such as Transformer models and Graph Neural Networks.