Research map / Multi-agent systems and games / Human-AI and goal-conditioned
Prediction Equilibrium for Dynamic Network Flows
Lukas Graf, Tobias Harks, Kostas Kollias, Michael Markl
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| use neural network predictor instead of constant predictor | dynamic prediction equilibrium routing · Synthetic network | average slowdown (%): 17.36 -> 0.00 | Improved |
| use linear regression predictor instead of constant predictor | dynamic prediction equilibrium routing · Anaheim I network | average slowdown (%): 22.11 -> 0.50 | Improved |
| use neural network predictor instead of constant predictor | dynamic prediction equilibrium routing · Anaheim II network | average slowdown (%): 6.68 -> 1.25 | Improved |
| use neural network predictor instead of constant predictor | dynamic prediction equilibrium routing · Nguyen network | average slowdown (%): 1.21 -> 0.35 | Improved |
| use linear predictor instead of constant predictor | dynamic prediction equilibrium routing · Nguyen network | average slowdown (%): 1.21 -> 2.56 | Hurt |
| use linear predictor instead of constant predictor | dynamic prediction equilibrium routing · Anaheim I network | average slowdown (%): 22.11 -> 14.67 | Improved |
| add regularization to linear predictor | linear predictor · Sioux Falls II network | average slowdown (%): 4.25 -> 4.99 | Hurt |
| use zero predictor instead of constant predictor | dynamic prediction equilibrium routing · Synthetic network | average slowdown (%): 17.36 -> 16.07 | Improved |
| use neural network predictor instead of linear regression predictor | ML predictors · Sioux Falls I network | average slowdown (%): 0.73 -> 0.06 | Improved |
| use neural network predictor instead of linear regression predictor | ML predictors · Nguyen network | average slowdown (%): 0.27 -> 0.35 | Hurt |
| use ML predictors instead of the other four predictors in larger networks | dynamic prediction equilibrium routing · larger real-world networks | average slowdown | Improved |
| use linear predictor instead of constant predictor (approximation quality) | DPE computation · Anaheim II network | epsilon approximation quality: 0.65 -> 6.57 | Hurt |
Closest papers by meaning
- Machine-Learned Prediction Equilibrium for Dynamic Traffic Assignment (AAAI 2022 · similarity 0.97)
- Information Design for Congestion Games with Unknown Demand (AAAI 2024 · similarity 0.85)
- Neural Combinatorial Optimization for Time Dependent Traveling Salesman Problem (NeurIPS 2025 · similarity 0.85)
- Fast Routing under Uncertainty: Adaptive Learning in Congestion Games via Exponential Weights (NeurIPS 2021 · similarity 0.85)
- WardropNet: Traffic Flow Predictions via Equilibrium-Augmented Learning (ICLR 2025 · similarity 0.85)
- Dynamic Latent Routing (NeurIPS 2026 · similarity 0.83)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
