4 / 5 Learning Theory & Phenomena

When is Black Box Shift Estimation identifiable?

Black Box Shift Estimation is used to infer target-domain class proportions from unlabelled target data. Which combination of assumptions is central to the original method?

A

P(x|y) is invariant across domains, and the black-box predictor's confusion matrix is invertible

B

The conditional label distribution P(y|x) is invariant, and the classifier is perfectly calibrated

C

The covariate marginal P(x) is invariant, and the classifier has zero training error

D

Both P(x|y) and P(y) remain invariant, while P(y|x) changes