Traffic Volume Prediction Based on Weather Parameters

With increase in number of vehicles, traffic congestion is on rise. It leads to a multitude of problems in urban areas. Commuters experience unpredictable travel times, resulting in delays, stress, and inefficiency. Inadequate planning and resource allocation hinder traffic management efforts, leading to congestion, increased accident risks, and reduced overall transportation system performance. A way to deal with these problems is predicting the flow of traffic and understanding patterns according to different urban areas. Existing research on traffic volume prediction has predominantly overlooked the influence of weather parameters, despite their potential impact. This paper fills this research gap by focusing on the integration of weather parameters into traffic volume prediction models and shedding light on the relationship among weather and traffic dynamics. The prediction algorithms employed in our study include Decision Trees, XGboost, Support Vector Machine and Random Forest are compared using prediction parameters including Mean Squared Error, Mean Absolute Error, Root Mean Squared Error, and R-square. After analyzing each Algorithm, the SVM algorithm arrives to be the best fit for predicting traffic volume on weather conditions.