On the Value of Target Data in Transfer Learning

Abstract

We aim to understand the value of additional labeled or unlabeled target data in transfer learning, for any given amount of source data; this is motivated by practical questions around minimizing sampling costs, whereby, target data is usually harder or costlier to acquire than source data, but can yield better accuracy. To this aim, we establish the first minimax-rates in terms of both source and target sample sizes, and show that performance limits are captured by new notions of discrepancy between source and target, which we refer to as transfer exponents.

Cited by in corpus (9)

• A survey on domain adaptation theory: learning bounds and theoretical guarantees
• Minimax Lower Bounds for Transfer Learning with Linear and One-hidden Layer Neural Networks
• Minimax Estimation for Personalized Federated Learning: An Alternative between FedAvg and Local Training?
• Learning Invariant Representations and Risks for Semi-supervised Domain Adaptation
• Weighted Training for Cross-Task Learning
• On Localized Discrepancy for Domain Adaptation
• Theoretical bounds on estimation error for meta-learning
• Characterizing and Understanding the Generalization Error of Transfer Learning with Gibbs Algorithm
• Learning to Transfer: A Foliated Theory