A review of deep learning models for traffic flow prediction in Autonomous Vehicles

More and more rapid development has been seen in the last decade in self-driving car technologies, mostly developments in the field of machine learning and deep learning. The goal of the paper is to review the latest state of the art in the area of automated driving by using machine learning technologies. For autonomous vehicle usage the estimation of traffic flows is necessary and they agree to make changes about their relevant artefacts (e.g., turn left or correct, travel straight, shift direction, stop or speed). Work on autonomous vehicles has been seen from the current paper on machine learning methods. Moreover, the non-linear dynamic relationship between spatial and temporal data obtained from the surroundings at the previously described adaptive decision-making periods by vehicles does not extend explicitly to current machine learning models in this context. Throughout this paper, we discussed the learning models for autonomous vehicle traffic flux prediction throughout order to equate such models with their applicability in contemporary intelligent transport systems. In comparison, the paper further addresses problems and possible recommendations for science.