Decision Theoretic Foundations for Conformal Prediction: Optimal Uncertainty Quantification for Risk-Averse Agents
Shayan Kiyani, George Pappas, Aaron Roth, Hamed Hassani · 2025
Read Decision Theoretic Foundations for Conformal Prediction: Optimal Uncertainty Quantification for Risk-Averse Agents by Shayan Kiyani, George Pappas, Aaron Roth, Hamed Hassani (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2502-02561.txt.
Description
Shayan Kiyani, George Pappas, Aaron Roth, and Hamed Hassani build decision theory for prediction sets used by cautious decision makers. Their 2025 paper starts from a practical question: how should uncertainty about a prediction be reported so it can guide action? The answer matters most in risk-sensitive fields like medicine.
They prove that prediction sets are the right form of uncertainty report for agents who care about value at risk. A simple worst-case policy, choosing the action that protects against the worst case in the set, is optimal for such agents. They also characterize the best sets exactly in theory and give a construction that works with finite data and no assumptions on the distribution. Their algorithm, built on these results, turns predictions into policies with provable backing.
Catalog record
| CREATOR | Shayan Kiyani, George Pappas, Aaron Roth, Hamed Hassani |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 92 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2502-02561