PROJECT TITLE :

A Survey on Modern Deep Neural Network for Traffic Prediction Trends, Methods and Challenges

ABSTRACT:

In this current era, traffic congestion has evolved into a major source of severe adverse effects on both the economy and the environment for urban areas all over the world. The ability to accurately forecast future traffic flows is one of the most effective strategies for reducing the negative effects of traffic congestion. Since its inception in the late 1970s, the research field of traffic prediction has undergone significant development ever since then. The majority of earlier studies make use of traditional statistical models like ARIMA and its various iterations. Researchers have only recently begun to concentrate their efforts on machine learning models due to the power and versatility of these models. As new theoretical and technological developments emerge, we enter the era of deep neural networks, which have gained popularity due to the sheer prediction power that can be attributed to the complex and deep structure. This popularity can be attributed to the fact that deep neural networks have complex and deep structures. Literature surveys of such methods are uncommon, despite the widespread use of deep neural network models in the field of traffic forecasting. [Citation needed] [Citation needed] An up-to-date survey of deep neural networks for the prediction of traffic is presented here as part of this body of work. We will first provide a detailed explanation of popular deep neural network architectures that are commonly used in the traffic flow prediction literatures, then we will categorize and describe the traffic flow prediction literatures themselves, then we will present an overview of the commonalities and differences among different works, and finally, we will provide a discussion regarding the challenges and future directions for this field.