Improving Urban Mobility in Dhaka: Machine Learning-Based Traffic Prediction

Urban traffic congestion is a pervasive issue with detrimental consequences for economies, environments, and quality of life. In rapidly urbanizing cities like Dhaka, the mismatch between population growth and infrastructure development exacerbates traffic congestion. To address this challenge, this paper proposes machine learning approaches to predict traffic congestion. Specifically, we employ Decision Tree algorithms, and K-Nearest Neighbors (KNN), trained on a real-life dataset, which is highly relevant and closely resembles the traffic scenarios of various locations in Dhaka over a two-month period. By considering factors such as vehicle counts, day of the week, and total traffic, the models classify traffic situations into congestion levels. The Decision Tree model, in particular, achieves an impressive accuracy of 99.58%, surpassing the performance of the KNN model in terms of both prediction and efficiency. These findings underscore the potential of Decision Tree models for real-time traffic congestion prediction and their contribution to the development of intelligent traffic management systems. By implementing such models, cities can mitigate congestion, optimize traffic flow, and enhance the overall quality of life for their residents.