Label-Wise uncertainty decomposition for Multi-label Classification by Maximizing Type II Likelihood

Minghao Li, Junjie Qiu, Weishi Shi
Proceedings of the 42nd Conference on Uncertainty in Artificial Intelligence, PMLR 337:3719-3737, 2026.

Abstract

Currently, the way deep learning models recognize uncertainty remains inconsistent with human perception. In multi-label classification, quantifying uncertainty at the label level presents challenges, as each label may exhibit distinct model confidence levels. Understanding and decomposing label-specific uncertainty is essential for interpreting model behavior and ensuring reliable predictions. We build a hierarchical {Bayesian} methodology for multi-label classification that leverages a Type {II} likelihood and Empirical {Bayes}. Then we estimate and decompose label-wise uncertainties by the bias-variance decomposition. Our approaches offer four main contributions: (1) Type {II} likelihood maximization is data likelihood centric; (2) it can decompose label-wise uncertainty into the model variance, the model bias and data noise; (3) our uncertainty represented by model bias is intuitively interpretable when combined with observational data; and (4) when applied to out-of-distribution ({OOD}) detection task, it achieves a 6.88% lower FPR95 score on NUS-WIDE.

Cite this Paper

BibTeX
@InProceedings{pmlr-v337-li26i, title = {Label-Wise uncertainty decomposition for Multi-label Classification by Maximizing Type {II} Likelihood}, author = {Li, Minghao and Qiu, Junjie and Shi, Weishi}, booktitle = {Proceedings of the 42nd Conference on Uncertainty in Artificial Intelligence}, pages = {3719--3737}, year = {2026}, editor = {Perković, Emilija and Malinsky, Daniel}, volume = {337}, series = {Proceedings of Machine Learning Research}, month = {17--21 Aug}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v337/main/assets/li26i/li26i.pdf}, url = {https://proceedings.mlr.press/v337/li26i.html}, abstract = {Currently, the way deep learning models recognize uncertainty remains inconsistent with human perception. In multi-label classification, quantifying uncertainty at the label level presents challenges, as each label may exhibit distinct model confidence levels. Understanding and decomposing label-specific uncertainty is essential for interpreting model behavior and ensuring reliable predictions. We build a hierarchical {Bayesian} methodology for multi-label classification that leverages a Type {II} likelihood and Empirical {Bayes}. Then we estimate and decompose label-wise uncertainties by the bias-variance decomposition. Our approaches offer four main contributions: (1) Type {II} likelihood maximization is data likelihood centric; (2) it can decompose label-wise uncertainty into the model variance, the model bias and data noise; (3) our uncertainty represented by model bias is intuitively interpretable when combined with observational data; and (4) when applied to out-of-distribution ({OOD}) detection task, it achieves a 6.88% lower FPR95 score on NUS-WIDE.} }
Endnote
%0 Conference Paper %T Label-Wise uncertainty decomposition for Multi-label Classification by Maximizing Type II Likelihood %A Minghao Li %A Junjie Qiu %A Weishi Shi %B Proceedings of the 42nd Conference on Uncertainty in Artificial Intelligence %C Proceedings of Machine Learning Research %D 2026 %E Emilija Perković %E Daniel Malinsky %F pmlr-v337-li26i %I PMLR %P 3719--3737 %U https://proceedings.mlr.press/v337/li26i.html %V 337 %X Currently, the way deep learning models recognize uncertainty remains inconsistent with human perception. In multi-label classification, quantifying uncertainty at the label level presents challenges, as each label may exhibit distinct model confidence levels. Understanding and decomposing label-specific uncertainty is essential for interpreting model behavior and ensuring reliable predictions. We build a hierarchical {Bayesian} methodology for multi-label classification that leverages a Type {II} likelihood and Empirical {Bayes}. Then we estimate and decompose label-wise uncertainties by the bias-variance decomposition. Our approaches offer four main contributions: (1) Type {II} likelihood maximization is data likelihood centric; (2) it can decompose label-wise uncertainty into the model variance, the model bias and data noise; (3) our uncertainty represented by model bias is intuitively interpretable when combined with observational data; and (4) when applied to out-of-distribution ({OOD}) detection task, it achieves a 6.88% lower FPR95 score on NUS-WIDE.
APA
Li, M., Qiu, J. & Shi, W.. (2026). Label-Wise uncertainty decomposition for Multi-label Classification by Maximizing Type II Likelihood. Proceedings of the 42nd Conference on Uncertainty in Artificial Intelligence, in Proceedings of Machine Learning Research 337:3719-3737 Available from https://proceedings.mlr.press/v337/li26i.html.

Related Material