Traffic Prediction Analysis (TPA) Using Machine Learning Methodologies

A traffic prediction system might help drivers in making better travel choices. Minimizing carbon emissions, reducing traffic congestion, and improving the effectiveness of traffic management to deliver such traffic flow data is the main aim of traffic flow prediction. Prediction of traffic is essential in a system of intelligent transportation. Effective traffic forecasting can help with route design, vehicle dispatching, and reducing traffic congestion. Due to its incredible capacity to predict traffic patterns, a branch of machine learning which is also called “deep learning,” has recently attracted a lot of attention. In this study, a comparative evaluation has been done to predict traffic flow based on deep learning. To prepare the stacked auto encoder model, basic traffic flow features are included. The goal of this study is to conduct a thorough screening for deep earning -based traffic forecast algorithms from multiple angles. This study recommends a deep learning method to predict urban traffic that combines data from twitter messages with traffic and weather data. The predictive model uses a deep Bi-directional long- short-term memory (LSTM) Stacked Auto Encoder (SAE) architecture for multi-step traffic flow prediction that was trained from tweets, traffic, and weather datasets. To support study in this area, it also gathers and arranges frequently available public datasets.