Pandora's Regret is a closed-form pairwise scoring rule derived from expected optimal search costs that elicits true probabilities and outperforms log loss, accuracy, and F1 at predicting diagnostic costs on MedMNIST models.
Convexity, classification, and risk bounds
8 Pith papers cite this work. Polarity classification is still indexing.
8
Pith papers citing it
representative citing papers
First approximate calibration results for discrete properties in multiclass settings via Lipschitz intermediaries for strongly orderable discrete properties.
A hybrid method of oracle-guided gradient descent and interval arithmetic generates increasingly tight certified lower bounds on the maximum satisfaction probability for stochastic constraints.
EEG measures of early cortical preconfiguration dynamics distinguish repetitive subconcussion patients from healthy controls and chronic TBI cases.
citing papers explorer
-
Pandora's Regret: A Proper Scoring Rule for Evaluating Sequential Search
Pandora's Regret is a closed-form pairwise scoring rule derived from expected optimal search costs that elicits true probabilities and outperforms log loss, accuracy, and F1 at predicting diagnostic costs on MedMNIST models.
-
Smoothed Elicitation Complexity for Approximate $\Gamma$-calibration of Discrete Classification Tasks
First approximate calibration results for discrete properties in multiclass settings via Lipschitz intermediaries for strongly orderable discrete properties.
-
Solving Stochastic Constraints by Oracle-based Gradient Descent and Interval Arithmetic
A hybrid method of oracle-guided gradient descent and interval arithmetic generates increasingly tight certified lower bounds on the maximum satisfaction probability for stochastic constraints.
-
Early Preconfiguration Failure: A Novel Predictor of the Repetitive Subconcussion
EEG measures of early cortical preconfiguration dynamics distinguish repetitive subconcussion patients from healthy controls and chronic TBI cases.
- Online Learning-to-Defer with Varying Experts
- Beyond Augmented-Action Surrogates for Multi-Expert Learning-to-Defer
- Adversarial Robustness in One-Stage Learning-to-Defer
- Optimal Query Allocation in Extractive QA with LLMs: A Learning-to-Defer Framework with Theoretical Guarantees