Online-LoRA: Task-free Online Continual Learning via Low Rank Adaptation
Catastrophic forgetting is a significant challenge in online continual learning (OCL),
especially for non-stationary data streams that do not have well-defined task boundaries …
especially for non-stationary data streams that do not have well-defined task boundaries …
Multi-Granularity Class Prototype Topology Distillation for Class-Incremental Source-Free Unsupervised Domain Adaptation
P Deng, J Zhang, X Sheng, C Yan, Y Sun, Y Fu… - arxiv preprint arxiv …, 2024 - arxiv.org
This paper explores the Class-Incremental Source-Free Unsupervised Domain Adaptation
(CI-SFUDA) problem, where the unlabeled target data come incrementally without access to …
(CI-SFUDA) problem, where the unlabeled target data come incrementally without access to …
Variable Forward Regularization to Replace Ridge in Online Linear Regression
J Wang, M Hu, N Li, PN Suganthan - openreview.net
Forward regularization (-F) with unsupervised knowledge was proposed to replace
canonical Ridge regularization (-R) in online linear learners, which achieves lower relative …
canonical Ridge regularization (-R) in online linear learners, which achieves lower relative …