The paper “Angular Regularization for Positive-Unlabeled Learning on the Hypersphere” received TMLR certification and was included in the TMLR weekly digest of 7 December 2025.
The method introduces AngularPU, a positive-unlabeled learning framework operating on the unit hypersphere. It uses cosine similarity, angular margins and an angular regularizer to improve separation in settings where only positive examples are explicitly labeled.
- Authors: Vasileios Sevetlidis, George Pavlidis, Antonios Gasteratos
- Venue: Transactions on Machine Learning Research (TMLR)
- TMLR digest date: 7 December 2025
- OpenReview: paper page
- TMLR announcement: weekly digest
You may also like
-
TMLR 2026: prescriptive SVD-inspired attention via spectral energy retention
-
Towards Heritage World Models: from digital twins to predictive heritage systems
-
ICASSP 2026: representation-diverse self-supervision for bioacoustic learning
-
AISTATS 2026: process-tensor tomography of SGD and training memory
-
ICPRAM 2026: two-stage angular alignment for positive-unlabeled learning