Table 1.
Literature summarized view.
Fig 1.
Proposed framework.
Table 2.
Dataset features and descriptions.
Fig 2.
Feature importance score.
Table 3.
Hyperparameter configuration for GNN-RMNet training.
Fig 3.
Correlation matrix of the features.
Fig 4.
Feature importance score.
Fig 5.
Driving speed distribution with common ranges and outliers.
Fig 6.
Geographical distribution of trips with normal and anomalous events.
Fig 7.
Geographical distribution of trips with normal and anomalous events.
Fig 8.
Distribution of trip duration.
Table 4.
Performance evaluation metrics comparison with methods.
Table 5.
Statistical analysis comparison of existing and proposed methods.
Fig 9.
Proposed GNN-RMNet training accuracy and loss.
Table 6.
Interpretability metrics for normal vs. anomalous trips.
Fig 10.
GNN-RMNet model’s parameter sensitivity analysis.