Adaptive Online Domain Incremental Continual Learning - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2022

Adaptive Online Domain Incremental Continual Learning

Résumé

Continual Learning (CL) problems pose significant challenges for Neural Network (NN)s. Online Domain Incremental Continual Learning (ODI-CL) refers to situations where the data distribution may change from one task to another. These changes can severely affect the learned model, focusing too much on previous data and failing to properly learn and represent new concepts. Conversely, if a model constantly forgets previously learned knowledge, it may be deemed too unstable and unsuitable. This work proposes Online Domain Incremental Pool (ODIP), a novel method to cope with catastrophic forgetting. ODIP also employs automatic concept drift detection and does not require task ids during training. ODIP maintains a pool of learners, freezing and storing the best one after training on each task. An additional Task Predictor (TP) is trained to select the most appropriate NN from the frozen pool for prediction. We compare ODIP against regularization methods and observe that it yields competitive predictive performance.
Fichier non déposé

Dates et versions

hal-04468400 , version 1 (20-02-2024)

Identifiants

Citer

Nuwan Gunasekara, Heitor Murilo Gomes, Albert Bifet, Bernhard Pfahringer. Adaptive Online Domain Incremental Continual Learning. Artificial Neural Networks and Machine Learning - ICANN 2022 - 31st International Conference on Artificial Neural Networks, Bristol, UK, September 6-9, 2022, Proceedings, Part I, Sep 2022, Bristol, United Kingdom. pp.491--502, ⟨10.1007/978-3-031-15919-0_41⟩. ⟨hal-04468400⟩
8 Consultations
0 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More