TKIL: Tangent Kernel Approach for Class Balanced Incremental Learning. (arXiv:2206.08492v1 [cs.LG])

Web: http://arxiv.org/abs/2206.08492

When learning new tasks in a sequential manner, deep neural networks tend to
forget tasks that they previously learned, a phenomenon called catastrophic
forgetting. Class incremental learning methods aim to address this problem by
keeping a memory of a few exemplars from previously learned tasks, and
distilling knowledge from them. However, existing methods struggle to balance
the performance across classes since they typically overfit the model to the
latest task. In our work, we propose to address these challenges with …

arxiv incremental kernel learning lg