This paper presents a state-of-the-art review of prediction models used to anticipate road traffic congestion. We present, on one hand, a classification and comparison of those models and, on the other hand, the results of our experimental tests. We provide a comprehensive list of prediction models such as Statistical models, Deep learning and hybrid models, Ensemble models, models based on big data, shallow learning models, naive methods, and traditional machine learning models like Support Vector Regression and Artificial Neural Networks (ANN). In addition to that, our paper also outlines various problems faced by road traffic users in congestion and the associated models designed to mitigate these issues. Experiments were performed to help users choose the appropriate algorithm, focusing on models such as Artificial Neural Networks (ANN), K-Nearest Neighbors (KNN), classification algorithms, as well as linear and nonlinear regression models.