Application of Different Machine Learning Methods to Predict Traffic Flow

Abstract

Nowadays, with the proliferation of automobiles, it's become more difficult to precisely anticipate traffic volumes. In recent years, traffic bottlenecks have become more widespread.  A dataset containing information about traffic volumes is used. Find the accuracy, mean absolute error, mean squared error, and Root Mean Squared Error (RMSE) for the anticipated traffic volume using Logistic Regression, Support vector Machines, random forest method, and Naive-Bayes machine learning algorithms.  Results from using Logistic Regression outperformed those from using Support Vector Machines, Random Forest and Naïve Bayes classifiers are verified using different metrics like MSE, MAE and Regression values.  Moreover, out of all the algorithms used, Logistic Regression obtained the lowest MSE of 1.1775e-30, lowest MAE of 7.9028e-16 and higher Regression Coefficient of 0.9999997.  Future plans for road construction and widening might be informed by the anticipated amount of traffic.