Robust Transfer Learning with Unreliable Source Data

TL;DR

This paper introduces a new measure called 'ambiguity level' for robust transfer learning, proposes the TAB model to improve classification performance, and demonstrates its effectiveness through theoretical bounds and empirical results.

Contribution

It presents the 'ambiguity level' as a novel metric and the 'Transfer Around Boundary' (TAB) model, enhancing robustness and efficiency in transfer learning tasks.

Findings

TAB improves classification accuracy and robustness.

Theoretical bounds on misclassification error are established.

Simulation results confirm the effectiveness of TAB.

Abstract

This paper addresses challenges in robust transfer learning stemming from ambiguity in Bayes classifiers and weak transferable signals between the target and source distribution. We introduce a novel quantity called the ''ambiguity level'' that measures the discrepancy between the target and source regression functions, propose a simple transfer learning procedure, and establish a general theorem that shows how this new quantity is related to the transferability of learning in terms of risk improvements. Our proposed ''Transfer Around Boundary'' (TAB) model, with a threshold balancing the performance of target and source data, is shown to be both efficient and robust, improving classification while avoiding negative transfer. Moreover, we demonstrate the effectiveness of the TAB model on non-parametric classification and logistic regression tasks, achieving upper bounds which are optimal up to logarithmic factors. Simulation studies lend further support to the effectiveness of TAB. We also provide simple approaches to bound the excess misclassification error without the need for specialized knowledge in transfer learning.