English

New metrics for analyzing continual learners

Machine Learning 2023-09-04 v1 Artificial Intelligence

Abstract

Deep neural networks have shown remarkable performance when trained on independent and identically distributed data from a fixed set of classes. However, in real-world scenarios, it can be desirable to train models on a continuous stream of data where multiple classification tasks are presented sequentially. This scenario, known as Continual Learning (CL) poses challenges to standard learning algorithms which struggle to maintain knowledge of old tasks while learning new ones. This stability-plasticity dilemma remains central to CL and multiple metrics have been proposed to adequately measure stability and plasticity separately. However, none considers the increasing difficulty of the classification task, which inherently results in performance loss for any model. In that sense, we analyze some limitations of current metrics and identify the presence of setup-induced forgetting. Therefore, we propose new metrics that account for the task's increasing difficulty. Through experiments on benchmark datasets, we demonstrate that our proposed metrics can provide new insights into the stability-plasticity trade-off achieved by models in the continual learning environment.

Keywords

Cite

@article{arxiv.2309.00462,
  title  = {New metrics for analyzing continual learners},
  author = {Nicolas Michel and Giovanni Chierchia and Romain Negrel and Jean-François Bercher and Toshihiko Yamasaki},
  journal= {arXiv preprint arXiv:2309.00462},
  year   = {2023}
}

Comments

6 pages, presented at MIRU 2023

R2 v1 2026-06-28T12:10:23.546Z