Label Shift Quantification with Robustness Guarantees via Distribution Feature Matching - Inria - Institut national de recherche en sciences et technologies du numérique Access content directly
Conference Papers Year : 2023

Label Shift Quantification with Robustness Guarantees via Distribution Feature Matching

Abstract

Quantification learning deals with the task of estimating the target label distribution under label shift. In this paper, we first present a unifying framework, distribution feature matching (DFM), that recovers as particular instances various estimators introduced in previous literature. We derive a general performance bound for DFM procedures, improving in several key aspects upon previous bounds derived in particular cases. We then extend this analysis to study robustness of DFM procedures in the misspecified setting under departure from the exact label shift hypothesis, in particular in the case of contamination of the target by an unknown distribution. These theoretical findings are confirmed by a detailed numerical study on simulated and real-world datasets. We also introduce an efficient, scalable and robust version of kernel-based DFM using the Random Fourier Feature principle.

Dates and versions

hal-04122205 , version 1 (08-06-2023)

Licence

Identifiers

Cite

Bastien Dussap, Gilles Blanchard, Badr-Eddine Chérief-Abdellatif. Label Shift Quantification with Robustness Guarantees via Distribution Feature Matching. ECML-PKDD 2023, 2023, Turin (IT), Italy. pp.69-85, ⟨10.1007/978-3-031-43424-2_5⟩. ⟨hal-04122205⟩
80 View
0 Download

Altmetric

Share

Gmail Mastodon Facebook X LinkedIn More