Utilizing Specialized Graph Partitioning and Adaptive GNNs: A Comparative Study for Large-Scale Traffic Forecasting
Bibliographic record
Abstract
As urban areas grow and transportation networks become increasingly complex, accurate large-scale traffic forecasting is essential for effective Intelligent Transportation Systems (ITS). Traditional statistical and machine learning approaches often fail to capture the spatial-temporal patterns in large-scale transportation data, while conventional deep learning methods struggle with non-Euclidean network structures. Graph Neural Networks (GNNs) address these challenges by modeling road networks as a graph, resulting in improved forecasting. This thesis builds on the state of the art in large-scale traffic prediction. This is done by integrating an Adaptive Graph Convolutional Recurrent Network (AGCRN) with a high-quality, domain-specific graph partitioning tool - the Buffoon-optimized Karlsruhe High Quality Partitioning (KaHIP). Building on the limitations observed in established models like the Diffusion Convolutional Recurrent Neural Network (DCRNN), AGCRN introduces node-adaptive parameters to more effectively learn localized, evolving traffic patterns. To overcome computational challenges associated with large networks, we employ specialized partitioning frameworks. While graph partitioners like METIS has been the standard choice, we demonstrate that the Buffoonoptimized KaHIP - framework produces higher quality partitions, resulting in higher forecasting accuracy. Tests conducted on a California highway network dataset show that the AGCRNKaHIP integration outperforms DCRNN-METIS and baseline statistical methods. The integration delivers lower prediction errors and more stable forecasts. These results highlight the value of using domain-specific partitioning strategies and adaptive GNN architectures for large-scale traffic forecasting.
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 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.002 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
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".