Research map / Learning theory and algorithms / Combinatorial and search optimization
Maximizing the Probability of Arriving on Time: A Practical Q-Learning Method
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 Q-learning with success-reward formulation instead of LET/mean-risk/cardinality baselines | path planning · Artificial grid, discrete deadlines | accuracy: about 95% -> 100% | Helped |
| use Q-learning with neural network instead of baselines on large real networks | path planning · Munich/Singapore/Beijing | accuracy: around 95% | Helped |
| use neural network function approximation for continuous deadlines instead of tabular discrete Q-learning | Q-learning method · real networks vs grid | accuracy | Hurt |
| use Q-learning instead of cardinality method (MILP) | path planning · Beijing network | computation time per routing: around 15 seconds -> 1.216 seconds | Helped |
| finer time resolution for training data (week to day to hour) (ablation) | time-dependent Q-learning · Munich | accuracy (%): 91.6 -> 96.1 (segment 8, week vs hour) | Helped |
| use only 1/3 of training samples (ablation) | Q-learning method · Beijing | accuracy and training time: accuracy drop around 1%, time reduced about 360 seconds | Mixed |
| reduce travel time data per link (ablation) | Q-learning method · Beijing | accuracy | No clear effect |
Closest papers by meaning
- Neural Combinatorial Optimization for Robust Routing Problem with Uncertain Travel Times (NeurIPS 2024 · similarity 0.83)
- Taking the Road Less Scheduled with Adaptive Polyak Steps (NeurIPS 2026 · similarity 0.83)
- WardropNet: Traffic Flow Predictions via Equilibrium-Augmented Learning (ICLR 2025 · similarity 0.83)
- GLOP: Learning Global Partition and Local Construction for Solving Large-Scale Routing Problems in Real-Time (AAAI 2024 · similarity 0.82)
- BO4Mob: Bayesian Optimization Benchmarks for High-Dimensional Urban Mobility Problem (NeurIPS 2025 · similarity 0.82)
- Improving Generalization of Neural Combinatorial Optimization for Vehicle Routing Problems via Test-Time Projection Learning (NeurIPS 2025 · similarity 0.82)
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.