The increasing population in urban areas has led to an increase in the number of vehicles. The growth of vehicles is not comparable with the growth of roads that caused traffic congestion. To solve the congestion problem, traffic conditions are predicted. The method used to predict traffic congestion is Support Vector Regression (SVR). The map-Matching algorithm is used to adjust vehicle position data that is not accurate, resulting in an accurate position. In this research, the performance of SVR was tested for congestion prediction. The test was conducted to determine the performance of the SVR model by calculating the Means Absolute Error (MAE) and Root Mean Squared Error (RMSE). K-fold cross-validation is used to show the accuracy of the SVR model on the dataset after being divided into k sections and FCE to determine the level of congestion. We conclude that the SVR has good performance for congestion prediction, this is indicated by a small error value and high accuracy value. The prediction results using SVR are the same as the actual condition.
Analysis Performance of Support Vector Regression (SVR) for Congestion Prediction
16.11.2022
338087 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
Travel-Time Prediction With Support Vector Regression
British Library Conference Proceedings | 2004
|Travel-time prediction with support vector regression
IEEE | 2004
|Travel Time Prediction with Support Vector Regression
British Library Conference Proceedings | 2003
|Travel time prediction with support vector regression
IEEE | 2003
|Prediction of Pavement Performance: Application of Support Vector Regression with Different Kernels
Transportation Research Record | 2016
|