BriefGPT.xyz
Mar, 2023
当今的连续学习算法有多高效?
How Efficient Are Today's Continual Learning Algorithms?
HTML
PDF
Md Yousuf Harun, Jhair Gallardo, Tyler L. Hayes, Christopher Kanan
TL;DR
本文研究了增量类学习的最新方法,并指出很多算法在计算、存储和内存方面效率极低,有些甚至比从头开始训练的方法更费时,作者认为在实际应用中,不能忽视这些算法的资源消耗,持续学习不仅仅只是为了减轻灾难性遗忘。
Abstract
supervised continual learning
involves updating a
deep neural network
(DNN) from an ever-growing stream of labeled data. While most work has focused on overcoming catastrophic forgetting, one of the major motivat
→