Easy Learning from Label Proportions
Preprint
Abstract
We consider the problem of Learning from Label Proportions (LLP), a weakly supervised classification setup where instances are grouped into "bags", and only the frequency of class labels at each bag is available. Albeit, the objective of the learner is to achieve low task loss at an individual instance level. Here we propose Easyllp: a flexible and simple-to-implement debiasing approach based on aggregate labels, which operates on arbitrary loss functions. Our technique allows us to accurately estimate the expected loss of an arbitrary model at an individual level. We showcase the flexibility of our approach by applying it to popular learning frameworks, like Empirical Risk Minimization (ERM) and Stochastic Gradient Descent (SGD) with provable guarantees on instance level performance. More concretely, we exhibit a variance reduction technique that makes the quality of LLP learning deteriorate only by a factor of k (k being bag size) in both ERM and SGD setups, as compared to full supervision. Finally, we validate our theoretical results on multiple datasets demonstrating our algorithm performs as well or better than previous LLP approaches in spite of its simplicity.Keywords
All Related Versions
This publication has 4 references indexed in Scilit:
- Weakly Supervised Classification For High Energy PhysicsJournal of Physics: Conference Series, 2018
- Learning from label proportions with pinball lossInternational Journal of Machine Learning and Cybernetics, 2017
- Learning from label proportions for SAR image classificationEURASIP Journal on Advances in Signal Processing, 2017
- Learning With Label Proportions via NPSVMIEEE Transactions on Cybernetics, 2016