aiqbench analysis
This research extends the Classification Without Labels (CWoLa) framework from binary cases to multiclass scenarios by leveraging posterior simplex geometry in mixture space. The authors demonstrate that training on unlabeled mixtures with varying class proportions allows a model to recover optimal latent discriminators and their fractions without requiring instance-level labels or prior knowledge of mixing matrices. Validated across image datasets like MNIST and astronomical data from Galaxy10, this approach offers a scalable solution for label-scarce domains where only weakly enriched samples are available.
Multiclass Classification without Labels via Posterior Simplex Geometry
arXiv:2607.24943v1 Announce Type: cross Abstract: In many classification problems, reliable instance-level labels are unavailable. However, it is often possible to construct weakly enriched unlabeled samples: datasets selected by different cuts, sources, populations, or experimental conditions that change latent class proportions without revealing them. Classification without Labels (CWoLa) shows that, in the binary case ($K=2$), a classifier trained to distinguish two impure mixtures with different class proportions can recover an optimal class discriminator without knowing the mixture proportions. We extend this principle to multiclass learning from several unlabeled mixtures ($K>2$), where the learner observes only mixture identity and neither latent class labels nor class-prior matrices. We prove that, for a multiclass mixture model, the Bayes-optimal mixture classifier $g^\star$ maps data points into a $(K-1)$-simplex embedded in mixture-posterior space. The $K$ vertices of this simplex are induced by the latent classes through the unknown mixing matrix. Leveraging this geometry, we propose prior-free procedures that train a standard classifier to distinguish mixture identities and then extract latent class structure using either post-hoc simplex fitting or a bottleneck architecture. Experiments on MNIST, CIFAR-10, and Galaxy10 DECaLS show that mixture identity alone can recover latent classes and their fractions in the mixture. By narrowing the gap between weakly supervised and fully supervised performance, we provide a mathematically grounded, scalable tool for multiclass discovery in label-scarce domains.
Abstract:In many classification problems, reliable instance-level labels are unavailable. However, it is often possible to construct weakly enriched unlabeled samples: datasets selected by different cuts, sources, populations, or experimental conditions that change latent class proportions without revealing them. Classification without Labels (CWoLa) shows that, in the binary case ($K=2$), a classifier trained to distinguish two impure mixtures with different class proportions can recover an optimal class discriminator without knowing the mixture proportions. We extend this principle to multiclass learning from several unlabeled mixtures ($K>2$), where the learner observes only mixture identity and neither latent class labels nor class-prior matrices. We prove that, for a multiclass mixture model, the Bayes-optimal mixture classifier $g^\star$ maps data points into a $(K-1)$-simplex embedded in mixture-posterior space. The $K$ vertices of this simplex are induced by the latent classes through the unknown mixing matrix. Leveraging this geometry, we propose prior-free procedures that train a standard classifier to distinguish mixture identities and then extract latent class structure using either post-hoc simplex fitting or a bottleneck architecture. Experiments on MNIST, CIFAR-10, and Galaxy10 DECaLS show that mixture identity alone can recover latent classes and their fractions in the mixture. By narrowing the gap between weakly supervised and fully supervised performance, we provide a mathematically grounded, scalable tool for multiclass discovery in label-scarce domains.
Machine Learning (cs.LG); Astrophysics of Galaxies (astro-ph.GA); Artificial Intelligence (cs.AI); Machine Learning (stat.ML)
arXiv:2607.24943 [cs.LG]
(or arXiv:2607.24943v1 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2607.24943 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 320 words
Related coverage
Focal Calibration Loss: Controlling Posterior Distortion in Deep Neural Classifiers
arXiv · Aug 26, 2026
Large Classification-Risk-Optional Label Acquisition
arXiv · Sep 9, 2026
Informative Label Missingness in Multiclass Classification Information Geometry and Excess Risk
arXiv · Sep 1, 2026
Simplex-to-Euclidean Bijection for Conjugate and Calibrated Multiclass Gaussian Process Classification
arXiv · Aug 31, 2026