SOTAVerified

Classifier calibration

Confidence calibration – the problem of predicting probability estimates representative of the true correctness likelihood – is important for classification models in many applications. The two common calibration metrics are Expected Calibration Error (ECE) and Maximum Calibration Error (MCE).

Papers

Recently AddedMost HypedMost ActiveNeeds VerificationMost Verified

Showing 21–29 of 29 papers

TitleStatusHype
Long-tailed Medical Diagnosis with Relation-aware Representation Learning and Iterative Classifier CalibrationCode0
Hidden Heterogeneity: When to Choose Similarity-Based CalibrationCode0
Expeditious Saliency-guided Mix-up through Random Gradient ThresholdingCode0
Enhancing Generalized Few-Shot Semantic Segmentation via Effective Knowledge TransferCode0
Class-wise and reduced calibration methodsCode0
No Fear of Heterogeneity: Classifier Calibration for Federated Learning with Non-IID DataCode0
Packed-Ensembles for Efficient Uncertainty EstimationCode0
Classifier Calibration: with application to threat scores in cybersecurityCode0
Accuracy-Preserving Calibration via Statistical Modeling on Probability SimplexCode0
Show:102550
← PrevPage 3 of 3Next →

No leaderboard results yet.