Towards the Generalization of Contrastive Self-Supervised Learning

Abstract

Recently, self-supervised learning has attracted great attention, since it only requires unlabeled data for model training. Contrastive learning is one popular method for self-supervised learning and has achieved promising empirical performance. However, the theoretical understanding of its generalization ability is still limited. To this end, we define a kind of $(\sigma ,\delta )$-measure to mathematically quantify the data augmentation, and then provide an upper bound of the downstream classification error rate based on the measure. It reveals that the generalization ability of contrastive self-supervised learning is related to three key factors: alignment of positive samples, divergence of class centers, and concentration of augmented data. The first two factors are properties of learned representations, while the third one is determined by pre-defined data augmentation. We further investigate two canonical contrastive losses, InfoNCE and cross-correlation, to show how they provably achieve the first two factors. Moreover, we conduct experiments to study the third factor, and observe a strong correlation between downstream performance and the concentration of augmented data.

Accepted by ICLR 2023

References in corpus (7)

• Bootstrap your own latent: A new approach to self-supervised Learning
• Barlow Twins: Self-Supervised Learning via Redundancy Reduction
• CLEAR: Contrastive Learning for Sentence Representation
• Theoretical Analysis of Self-Training with Deep Networks on Unlabeled Data
• Predicting What You Already Know Helps: Provable Self-Supervised Learning
• ConSERT: A Contrastive Framework for Self-Supervised Sentence Representation Transfer
• Can Pretext-Based Self-Supervised Learning Be Boosted by Downstream Data? A Theoretical Analysis

Cited by in corpus (1)

• Self-supervised remote sensing feature learning: Learning Paradigms, Challenges, and Future Works