The short-term prediction of traffic parameters: a review of parametric and nonparametric approaches

Over the past few decades, there has been a lot of interest in short-term traffic prediction which is a crucial part of transportation systems. One of the major considerations that travelers evaluate when making travel plans is information regarding the near future. Transportation planners have investigated a number of methods to produce more reliable travel time predictions in the future. However, there have not been enough in-depth and comprehensive surveys in this area yet. In this paper, a comprehensive review of the literature has been conducted, and various traffic parameter prediction algorithms have been investigated. The methods can be divided into two main groups: parametric, and nonparametric. Parametric models are those that require the specification of some parameters before they can be used to make predictions. In contrast to nonparametric procedures, parametric approaches have a distribution with a defined number of parameters. The parametric approaches are related to statistical methods like time-series, while the nonparametric approaches are related to machine learning methods like neural networks. Predicting flow, volume, speed, density, and occupancy are the main emphasis of the majority of the literature. A detailed methodology is first presented for each of the techniques mentioned in this article, and then the outcomes of numerous articles that have used the technique are discussed.