Latent-Topology Graph State-Space Model (LT-GSSM) for Robust Traffic Fore-Casting
Bibliographic record
Abstract
Accurate traffic forecasting remains challenging when sensor data are noisy, incomplete, or non-stationary. Recent advances in spatio-temporal learning have combined Graph Neural Networks (GNNs) with recurrent, convolutional, or attention mechanisms to capture spatio-temporal dependencies. However, most existing approaches remain largely deterministic and rely on fixed or pre-learned adjacency matrices, limiting their adaptability when network structures evolve or sensor reliability varies. Some methods further stack multiple adjacency matrices to represent complex spatial relations, yet still lack explicit mechanisms to model uncertainty, resulting in reduced robustness under degraded data conditions. This work introduces the Latent Topology Graph State-Space Model (LT-GSSM), a probabilistic framework designed to enhance robustness and adaptability in traffic forecasting. LT-GSSM represents the road network as a latent dynamic graph whose structure evolves over-time through dynamic adjacency learning based on past hidden states and observations, enabling the model to capture evolving spatial correlations such as congestion propagation. Temporal dependencies are modelled by a nonlinear state-space function implemented with a Temporal Convolutional Network (TCN), which captures long-range temporal patterns without recurrence. The probabilistic state-space formulation explicitly represents sensor noise and handles missing data through probabilistic estimation inspired by Kalman filtering. By jointly integrating dynamic graph learning, explicit noise modelling, and nonlinear temporal transitions, LT-GSSM achieves greater stability and resilience to data uncertainty. Experiments on SUMO simulations and real-world PeMS datasets show that LT-GSSM consistently outperforms static and adaptive-graph models, providing a strong foundation for robust spatio-temporal forecasting under uncertain conditions.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.003 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.002 | 0.001 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.002 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".