Third-party cyber evaluations involving OpenAI models
openai.com·13h ago
AI Research
arXiv CS.AI·Jun 15·5 min read
Learning Urban Access Costs from Origin-Destination Flows via Inverse Optimal Transport
TL;DR
Urban planners often lack insight into the costs influencing household choices for services like schools. This study employs inverse optimal transport techniques, specifically a distance-banded model and a neural cost model, to analyze school enrollment flows in the Philippines.
✦ Why It Matters
Engineers can leverage this methodology to enhance urban planning and service accessibility through data-driven insights.
Key Takeaways
- Utilize inverse optimal transport for urban planning.
- Analyze origin-destination data for service allocation.
- Estimate perceived travel costs to inform subsidies.
- Apply findings to improve educational access in cities.
How It Works
The framework uses two models to recover latent costs from observed school enrollment flows. The distance-banded model incorporates a subsidy term, while the neural cost model leverages a differentiable Sinkhorn algorithm to optimize the transport plan.
This dual approach allows for a comprehensive understanding of how subsidies influence travel costs and choices.