Transfer Learning for Evolving Domains
Abstract
Transfer learning explores how to leverage knowledge from various tasks or domains (sources) to enhance predictive performance in related tasks or domains (targets). Typically, transfer learning research is segmented into several isolated sub-areas (such as domain generalisation, domain adaptation, or multi-domain learning), each making distinct assumptions about target data availability, namely how much data and how many labels are available at training time. However, in many real-world applications, data availability is not fixed but evolves over time, as instances and labels are progressively collected from a new domain. Each of the classical settings then describes only a snapshot of a trajectory that a deployed system must traverse in full. We formalise this trajectory as a transfer learning problem in its own right, Transfer Learning for Evolving Domains (TrED), specified by a data availability process fixed by the environment, a learning protocol that the method is free to choose, and an evaluation criterion that scores the whole trajectory of models rather than a single one. Within this formalism, the classical settings are recovered as regimes that a learner may pass through, rather than as separate problems that TrED concatenates. We then examine the transfer learning literature to identify mechanisms that are promising building blocks for a solution, and find that most methods are tailored to a single regime and that even the strongest existing candidates do not yet optimise the whole trajectory. We argue that TrED is a well-posed and unsolved problem, and an important direction for future research.
Related
- When Risk Factors Compete: A Machine Learning Approach to Identifying Antecedents of Adolescent Delinquency Across Domains, Development, and Gender
- Transfer learning in bayesian optimization for aircraft design
- A Method for Exact Enforcement of Boundary Conditions on General Curved Triangular Domains: Algorithm and Application to Scientific Machine Learning
- Transfer Learning with Heterogeneous Feature Spaces in Linear Regression
- Direct Knowledge Transfer for Offline Q-Learning
BibTeX
@misc{pereira2026transfer,
title = {{Transfer Learning for Evolving Domains}},
author = {Ricardo Ribeiro Pereira and Jacopo Bono and Hugo Ferreira and Pedro Ribeiro and Pedro Saleiro and Pedro Bizarro and Carlos Soares},
howpublished = {arXiv},
year = {2026},
month = sep,
doi = {10.48550/arxiv.2609.13039},
eprint = {2609.13039},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2609.13039},
note = {CC BY-NC-ND 4.0},
}
RIS
TY - UNPB
TI - Transfer Learning for Evolving Domains
AU - Ricardo Ribeiro Pereira
AU - Jacopo Bono
AU - Hugo Ferreira
AU - Pedro Ribeiro
AU - Pedro Saleiro
AU - Pedro Bizarro
AU - Carlos Soares
PY - 2026
DA - 2026/09/11
PB - arXiv
DO - 10.48550/arxiv.2609.13039
UR - https://arxiv.org/abs/2609.13039
AB - Transfer learning explores how to leverage knowledge from various tasks or domains (sources) to enhance predictive performance in related tasks or domains (targets). Typically, transfer learning research is segmented into several isolated sub-areas (such as domain generalisation, domain adaptation, or multi-domain learning), each making distinct assumptions about target data availability, namely how much data and how many labels are available at training time. However, in many real-world applications, data availability is not fixed but evolves over time, as instances and labels are progressively collected from a new domain. Each of the classical settings then describes only a snapshot of a trajectory that a deployed system must traverse in full. We formalise this trajectory as a transfer learning problem in its own right, Transfer Learning for Evolving Domains (TrED), specified by a data availability process fixed by the environment, a learning protocol that the method is free to choose, and an evaluation criterion that scores the whole trajectory of models rather than a single one. Within this formalism, the classical settings are recovered as regimes that a learner may pass through, rather than as separate problems that TrED concatenates. We then examine the transfer learning literature to identify mechanisms that are promising building blocks for a solution, and find that most methods are tailored to a single regime and that even the strongest existing candidates do not yet optimise the whole trajectory. We argue that TrED is a well-posed and unsolved problem, and an important direction for future research.
ER -