Answers are grounded in this paper and cite their sources.
One of the foremost challenges facing rapid urban areas is traffics congestions.In many of these areas, traffic is unmonitored and management is uncoordinated.The study introduces Fuzzy Graph Neural Networks (FGNNs) to the Intelligent Traffic Advisory Model (ITAM) and uses it to classify congestion types and suggest context-aware congestion resolutions in Port Harcourt, Nigeria.The work is based on the outcome of a design science innovation project.The project considers uncertainty in both the nature and level of traffic indicators, and connects road segments with traffic indicators using a fuzzy connected graph.The model represents input features such as traffic volume, weather, recorded traffic, speed and hazards, and output features like congestion.For indicators with uncertainty and fuzziness, Fuzzy logic is used to differentiate among low, moderate and severe congestion.For dependencies among road segments, Graph Neural Networks (GNNs) is employed.The source study implements Python and related ML frameworks, powered by a web-based front-end and a data pipeline integrated with a database.According to the study, training and validation accuracies converged to 91.2% and 92.7%, respectively, after 20 epochs.The model is illustrated as a means to discern congestion categories and suggests alternative routes.One of the evaluation sections reported 98% accuracy, 91% precision, 96% recall and 99% F1 score.The other section reports the average of the above scores as 92%, 90%, 88%, and 89%, respectively.Discrepancies in the above reported results are to be verified against the original experimental records prior to publishing in a journal.The study outlines an FGNN framework to specify the context and provides a detailed methodology for the evaluation and validation of models and recommendations for improving validation that includes available data, comparative baseline models, assessment of uncertainty, and preparation for field operation.
No takes yet. Share an insight, caveat, or question.
Okpobo et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: