Support Vector Regression Based Traffic Prediction Machine Learning Model*

To combat the rising issue of urban traffic congestion, optimization of traffic light timings at intersections is crucial. Traditional methods show stagnancy in managing traffic flow dynamically. This study introduces Support Vector Regression (SVR) as an unconventional solution for traffic signal optimization. The model predicts the traffic conditions and alters the green light to turn on and off accordingly. Utilization of a comprehensive dataset consisting of vehicle distributions and critical flow ratios have been implemented to training a machine learning model. The statistics and performance is checked by usage of Mean Squared Error, R Squared, Mean Absolute Error demonstrating that SVR can effectively enhance traffic signal control. This approach pledges significant improvements in reduction in congestion and streamlined traffic flow. Our findings majorly highlight the capability of the SVR Machine Learning model in improving urban traffic management systems and call for further enhancement and collaboration with experts to address dynamic traffic issues.