Factorizable joint shift revisited
Foundations of Data Science · 2026 · 0-0
DOI: 10.3934/fods.2026019 ↗Kurz gefasst
Fachartikel in der Zeitschrift Foundations of Data Science, 2026. Verfasst von Dirk Tasche. Die Arbeit nennt 46 Literaturangaben.
Abstract (Originalsprache)
Factorizable joint shift (FJS) represents a type of distribution shift (or dataset shift) that comprises both covariate and label shift. Recently, it has been observed that FJS actually arises from consecutive label and covariate (or vice versa) shifts. Research into FJS so far has been confined mostly to the case of categorical labels. We propose a framework for analysing distribution shift in the case of a general label space, thus covering both classification and regression models. Based on the framework, we generalise existing results on FJS to general label spaces and present and analyse a related extension to label distribution estimation of the expectation maximisation (EM) algorithm for class prior probabilities. We also take a fresh look at generalized label shift (GLS) in the case of a general label space.
Zitieren
APA
Tasche, D. (2026). Factorizable joint shift revisited. Foundations of Data Science, 0(0), 0-0. https://doi.org/10.3934/fods.2026019
Vancouver
Tasche D. Factorizable joint shift revisited. Foundations of Data Science. 2026;0(0):0-0. doi:10.3934/fods.2026019
BibTeX
@article{Tasche2026,
author = {Tasche, Dirk},
title = {Factorizable joint shift revisited},
journal = {Foundations of Data Science},
year = {2026},
volume = {0},
number = {0},
pages = {0--0},
doi = {10.3934/fods.2026019}
}
EUVIMED