Strength from Weakness: Fast Learning Using Weak Supervision

Abstract

We study generalization properties of weakly supervised learning. That is, learning where only a few "strong" labels (the actual target of our prediction) are present but many more "weak" labels are available. In particular, we show that having access to weak labels can significantly accelerate the learning rate for the strong task to the fast rate of $\mathcal{O}(\nicefrac1n)$, where $n$ denotes the number of strongly labeled data points. This acceleration can happen even if by itself the strongly labeled data admits only the slower $\mathcal{O}(\nicefrac{1}{\sqrt{n}})$ rate. The actual acceleration depends continuously on the number of weak labels available, and on the relation between the two tasks. Our theoretical results are reflected empirically across a range of tasks and illustrate how weak labels speed up learning on the strong task.

21 pages, 8 figures

References in corpus (7)

• PyTorch: An Imperative Style, High-Performance Deep Learning Library
• Deep Learning Recommendation Model for Personalization and Recommendation Systems
• WebVision Database: Visual Learning and Understanding from Web Data
• What makes ImageNet good for transfer learning?
• A Theoretical Analysis of Contrastive Unsupervised Representation Learning
• Self-supervised Visual Feature Learning with Deep Neural Networks: A Survey
• Fast learning rates in statistical inference through aggregation