Reinforced Domain Selection for Continuous Domain Adaptation

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Liu, Hanbing, Tang, Huaze, Wu, Yanru, Li, Yang, Zhang, Xiao-Ping
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866915548767453184
author Liu, Hanbing
Tang, Huaze
Wu, Yanru
Li, Yang
Zhang, Xiao-Ping
author_facet Liu, Hanbing
Tang, Huaze
Wu, Yanru
Li, Yang
Zhang, Xiao-Ping
contents Continuous Domain Adaptation (CDA) effectively bridges significant domain shifts by progressively adapting from the source domain through intermediate domains to the target domain. However, selecting intermediate domains without explicit metadata remains a substantial challenge that has not been extensively explored in existing studies. To tackle this issue, we propose a novel framework that combines reinforcement learning with feature disentanglement to conduct domain path selection in an unsupervised CDA setting. Our approach introduces an innovative unsupervised reward mechanism that leverages the distances between latent domain embeddings to facilitate the identification of optimal transfer paths. Furthermore, by disentangling features, our method facilitates the calculation of unsupervised rewards using domain-specific features and promotes domain adaptation by aligning domain-invariant features. This integrated strategy is designed to simultaneously optimize transfer paths and target task performance, enhancing the effectiveness of domain adaptation processes. Extensive empirical evaluations on datasets such as Rotated MNIST and ADNI demonstrate substantial improvements in prediction accuracy and domain selection efficiency, establishing our method's superiority over traditional CDA approaches.
format Preprint
id arxiv_https___arxiv_org_abs_2510_10530
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Reinforced Domain Selection for Continuous Domain Adaptation
Liu, Hanbing
Tang, Huaze
Wu, Yanru
Li, Yang
Zhang, Xiao-Ping
Machine Learning
Continuous Domain Adaptation (CDA) effectively bridges significant domain shifts by progressively adapting from the source domain through intermediate domains to the target domain. However, selecting intermediate domains without explicit metadata remains a substantial challenge that has not been extensively explored in existing studies. To tackle this issue, we propose a novel framework that combines reinforcement learning with feature disentanglement to conduct domain path selection in an unsupervised CDA setting. Our approach introduces an innovative unsupervised reward mechanism that leverages the distances between latent domain embeddings to facilitate the identification of optimal transfer paths. Furthermore, by disentangling features, our method facilitates the calculation of unsupervised rewards using domain-specific features and promotes domain adaptation by aligning domain-invariant features. This integrated strategy is designed to simultaneously optimize transfer paths and target task performance, enhancing the effectiveness of domain adaptation processes. Extensive empirical evaluations on datasets such as Rotated MNIST and ADNI demonstrate substantial improvements in prediction accuracy and domain selection efficiency, establishing our method's superiority over traditional CDA approaches.
title Reinforced Domain Selection for Continuous Domain Adaptation
topic Machine Learning
url https://arxiv.org/abs/2510.10530