A Systematic Review of Deep Learning Architectures in Traffic Congestion Prediction

The traffic data is increasing day by day. Data produced by traffic sensors, GPS equipped vehicles; loop detectors are used in intelligent transport systems. These data may include traffic volume, speed, density, local events, geospatial data, accident information and weather data. The collected data can be used in predicting the traffic flow, speed, congestion prediction. The more accurate prediction of congestion can increase the capacity of the road. In this paper, a review about congestion prediction is done with deep learning algorithms, metrics for classification and congestion.