Heterogeneous Label Shift: Theory and Algorithm

Chao Xu, Xijia Tang, Chenping Hou
Proceedings of the 42nd International Conference on Machine Learning, PMLR 267:69705-69724, 2025.

Abstract

In open-environment applications, data are often collected from heterogeneous modalities with distinct encodings, resulting in feature space heterogeneity. This heterogeneity inherently induces label shift, making cross-modal knowledge transfer particularly challenging when the source and target data exhibit simultaneous heterogeneous feature spaces and shifted label distributions. Existing studies address only partial aspects of this issue, leaving the broader problem unresolved. To bridge this gap, we introduce a new concept of Heterogeneous Label Shift (HLS), targeting this critical but underexplored challenge. We first analyze the impact of heterogeneous feature spaces and label distribution shifts on model generalization and introduce a novel error decomposition theorem. Based on these insights, we propose a bound minimization HLS framework that decouples and tackles feature heterogeneity and label shift accordingly. Extensive experiments on various benchmarks for cross-modal classification validate the effectiveness and practical relevance of the proposed approach.

Cite this Paper

BibTeX
@InProceedings{pmlr-v267-xu25ab, title = {Heterogeneous Label Shift: Theory and Algorithm}, author = {Xu, Chao and Tang, Xijia and Hou, Chenping}, booktitle = {Proceedings of the 42nd International Conference on Machine Learning}, pages = {69705--69724}, year = {2025}, editor = {Singh, Aarti and Fazel, Maryam and Hsu, Daniel and Lacoste-Julien, Simon and Berkenkamp, Felix and Maharaj, Tegan and Wagstaff, Kiri and Zhu, Jerry}, volume = {267}, series = {Proceedings of Machine Learning Research}, month = {13--19 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v267/main/assets/xu25ab/xu25ab.pdf}, url = {https://proceedings.mlr.press/v267/xu25ab.html}, abstract = {In open-environment applications, data are often collected from heterogeneous modalities with distinct encodings, resulting in feature space heterogeneity. This heterogeneity inherently induces label shift, making cross-modal knowledge transfer particularly challenging when the source and target data exhibit simultaneous heterogeneous feature spaces and shifted label distributions. Existing studies address only partial aspects of this issue, leaving the broader problem unresolved. To bridge this gap, we introduce a new concept of Heterogeneous Label Shift (HLS), targeting this critical but underexplored challenge. We first analyze the impact of heterogeneous feature spaces and label distribution shifts on model generalization and introduce a novel error decomposition theorem. Based on these insights, we propose a bound minimization HLS framework that decouples and tackles feature heterogeneity and label shift accordingly. Extensive experiments on various benchmarks for cross-modal classification validate the effectiveness and practical relevance of the proposed approach.} }
Endnote
%0 Conference Paper %T Heterogeneous Label Shift: Theory and Algorithm %A Chao Xu %A Xijia Tang %A Chenping Hou %B Proceedings of the 42nd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2025 %E Aarti Singh %E Maryam Fazel %E Daniel Hsu %E Simon Lacoste-Julien %E Felix Berkenkamp %E Tegan Maharaj %E Kiri Wagstaff %E Jerry Zhu %F pmlr-v267-xu25ab %I PMLR %P 69705--69724 %U https://proceedings.mlr.press/v267/xu25ab.html %V 267 %X In open-environment applications, data are often collected from heterogeneous modalities with distinct encodings, resulting in feature space heterogeneity. This heterogeneity inherently induces label shift, making cross-modal knowledge transfer particularly challenging when the source and target data exhibit simultaneous heterogeneous feature spaces and shifted label distributions. Existing studies address only partial aspects of this issue, leaving the broader problem unresolved. To bridge this gap, we introduce a new concept of Heterogeneous Label Shift (HLS), targeting this critical but underexplored challenge. We first analyze the impact of heterogeneous feature spaces and label distribution shifts on model generalization and introduce a novel error decomposition theorem. Based on these insights, we propose a bound minimization HLS framework that decouples and tackles feature heterogeneity and label shift accordingly. Extensive experiments on various benchmarks for cross-modal classification validate the effectiveness and practical relevance of the proposed approach.
APA
Xu, C., Tang, X. & Hou, C.. (2025). Heterogeneous Label Shift: Theory and Algorithm. Proceedings of the 42nd International Conference on Machine Learning, in Proceedings of Machine Learning Research 267:69705-69724 Available from https://proceedings.mlr.press/v267/xu25ab.html.

Related Material