English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Modern image classifiers are very accurate, but the predictions come withoutuncertainty estimates. Conformal predictors provide uncertainty estimates bycomputing a set of classes containing the correct class with a user-specifiedprobability based on the classifier's probability estimates. To provide suchsets, conformal predictors often estimate a cutoff threshold for theprobability estimates based on a calibration set. Conformal predictorsguarantee reliability only when the calibration set is from the samedistribution as the test set. Therefore, conformal predictors need to berecalibrated for new distributions. However, in practice, labeled data from newdistributions is rarely available, making calibration infeasible. In this work,we consider the problem of predicting the cutoff threshold for a newdistribution based on unlabeled examples. While it is impossible in general toguarantee reliability when calibrating based on unlabeled examples, we proposea method that provides excellent uncertainty estimates under naturaldistribution shifts, and provably works for a specific model of a distributionshift.

Test-time Recalibration of Conformal Predictors Under Distribution Shift Based on Unlabeled Examples