To search, Click below search items. |
| All Published Papers Search Service |
|
|
|
Title |
Graph-Based Spatio-Temporal Deep Learning for Enhanced Short-Term Traffic Prediction |
|
Author |
Madan Singh, Azween Abdullah |
| Citation |
Vol. 26 No. 3 pp. 182-194 |
|
Abstract |
Traffic forecasting in urban areas is a pivotal component of intelligent transportation systems (ITS). The paper proposes a novel Spatio - temporal predictive modelling methodology for traffic flow. An advanced deep learning model has been developed that integrate graph-based spatial learning with sequence-based temporal learning to capture complex road network interactions and traffic dynamics. The model leverage graph convolutional neural networks (GCN) and long short-term memory (LSTM) units to learn spatial dependencies between road segments and temporal patterns of traffic flow. The model has been evaluated on real-world traffic datasets and demonstrates improved accuracy and efficiency in short-term traffic prediction compared to state-of-the-art methods. The results show that the approach can better anticipate traffic congestion hot-spots and flow variations, contributing to proactive traffic management, reduced congestion, and enhanced urban mobility. A comprehensive analysis has been performed for related spatio-temporal prediction techniques. The proposed methodology offers a promising direction for intelligent traffic forecasting in urban settings. |
|
Keywords |
Intelligent Transportation Systems (ITS); Traffic Prediction; Spatio-Temporal Modelling; Graph Neural Networks; Deep Learning; Urban Mobility Forecasting. |
|
URL |
