Back to Search Start Over

Phase Transitions in Transfer Learning for High-Dimensional Perceptrons

Authors :
Dhifallah, Oussama
Lu, Yue M.
Publication Year :
2021

Abstract

Transfer learning seeks to improve the generalization performance of a target task by exploiting the knowledge learned from a related source task. Central questions include deciding what information one should transfer and when transfer can be beneficial. The latter question is related to the so-called negative transfer phenomenon, where the transferred source information actually reduces the generalization performance of the target task. This happens when the two tasks are sufficiently dissimilar. In this paper, we present a theoretical analysis of transfer learning by studying a pair of related perceptron learning tasks. Despite the simplicity of our model, it reproduces several key phenomena observed in practice. Specifically, our asymptotic analysis reveals a phase transition from negative transfer to positive transfer as the similarity of the two tasks moves past a well-defined threshold.

Details

Database :
arXiv
Publication Type :
Report
Accession number :
edsarx.2101.01918
Document Type :
Working Paper
Full Text :
https://doi.org/10.3390/e23040400