Uncertainty-based Continual Learning with Adaptive Regularization
- Authors
- Ahn, H[Ahn, Hongjoon]; Cha, S[Cha, Sungmin]; Lee, D[Lee, Donggyu]; Moon, T[Moon, Taesup]
- Issue Date
- 2019
- Publisher
- NEURAL INFORMATION PROCESSING SYSTEMS (NIPS)
- Citation
- ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS 32 (NIPS 2019), v.32
- Indexed
- OTHER
- Journal Title
- ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS 32 (NIPS 2019)
- Volume
- 32
- URI
- https://scholarworks.bwise.kr/skku/handle/2021.sw.skku/95292
- ISSN
- 1049-5258
- Abstract
- We introduce a new neural network-based continual learning algorithm, dubbed as Uncertainty-regularized Continual Learning (UCL), which builds on traditional Bayesian online learning framework with variational inference. We focus on two significant drawbacks of the recently proposed regularization-based methods: a) considerable additional memory cost for determining the per-weight regularization strengths and b) the absence of gracefully forgetting scheme, which can prevent performance degradation in learning new tasks. In this paper, we show UCL can solve these two problems by introducing a fresh interpretation on the Kullback-Leibler (KL) divergence term of the variational lower bound for Gaussian mean-field approximation. Based on the interpretation, we propose the notion of node-wise uncertainty, which drastically reduces the number of additional parameters for implementing per-weight regularization. Moreover, we devise two additional regularization terms that enforce stability by freezing important parameters for past tasks and allow plasticity by controlling the actively learning parameters for a new task. Through extensive experiments, we show UCL convincingly outperforms most of recent state-of-the-art baselines not only on popular supervised learning benchmarks, but also on challenging lifelong reinforcement learning tasks.
- Files in This Item
- There are no files associated with this item.
- Appears in
Collections - Information and Communication Engineering > School of Electronic and Electrical Engineering > 1. Journal Articles
![qrcode](https://api.qrserver.com/v1/create-qr-code/?size=55x55&data=https://scholarworks.bwise.kr/skku/handle/2021.sw.skku/95292)
Items in ScholarWorks are protected by copyright, with all rights reserved, unless otherwise indicated.