Cargando…
Phase Transitions in Transfer Learning for High-Dimensional Perceptrons
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-cal...
Autores principales: | , |
---|---|
Formato: | Online Artículo Texto |
Lenguaje: | English |
Publicado: |
MDPI
2021
|
Materias: | |
Acceso en línea: | https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8066313/ https://www.ncbi.nlm.nih.gov/pubmed/33801733 http://dx.doi.org/10.3390/e23040400 |
_version_ | 1783682545453367296 |
---|---|
author | Dhifallah, Oussama Lu, Yue M. |
author_facet | Dhifallah, Oussama Lu, Yue M. |
author_sort | Dhifallah, Oussama |
collection | PubMed |
description | 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. |
format | Online Article Text |
id | pubmed-8066313 |
institution | National Center for Biotechnology Information |
language | English |
publishDate | 2021 |
publisher | MDPI |
record_format | MEDLINE/PubMed |
spelling | pubmed-80663132021-04-25 Phase Transitions in Transfer Learning for High-Dimensional Perceptrons Dhifallah, Oussama Lu, Yue M. Entropy (Basel) Article 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. MDPI 2021-03-27 /pmc/articles/PMC8066313/ /pubmed/33801733 http://dx.doi.org/10.3390/e23040400 Text en © 2021 by the authors. https://creativecommons.org/licenses/by/4.0/Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/ (https://creativecommons.org/licenses/by/4.0/) ). |
spellingShingle | Article Dhifallah, Oussama Lu, Yue M. Phase Transitions in Transfer Learning for High-Dimensional Perceptrons |
title | Phase Transitions in Transfer Learning for High-Dimensional Perceptrons |
title_full | Phase Transitions in Transfer Learning for High-Dimensional Perceptrons |
title_fullStr | Phase Transitions in Transfer Learning for High-Dimensional Perceptrons |
title_full_unstemmed | Phase Transitions in Transfer Learning for High-Dimensional Perceptrons |
title_short | Phase Transitions in Transfer Learning for High-Dimensional Perceptrons |
title_sort | phase transitions in transfer learning for high-dimensional perceptrons |
topic | Article |
url | https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8066313/ https://www.ncbi.nlm.nih.gov/pubmed/33801733 http://dx.doi.org/10.3390/e23040400 |
work_keys_str_mv | AT dhifallahoussama phasetransitionsintransferlearningforhighdimensionalperceptrons AT luyuem phasetransitionsintransferlearningforhighdimensionalperceptrons |