Pruning neural networks using multi-armed bandits

Ameen, SA; Vadera, S

Abstract

The successful application of deep learning has led to increasing expectations
of their use in embedded systems. This in turn, has created the need to find
ways of reducing the size of neural networks. Decreasing the size of a neural
network requires deciding which weights should be removed without compromising
accuracy, which is analogous to the kind of problems addressed by multi-arm
bandits. Hence, this paper explores the use of multi-armed bandits for reducing
the number of parameters of a neural network. Different multi-armed bandit
algorithms, namely e-greedy, win-stay, lose-shift, UCB1, KL-UCB, BayesUCB,
UGapEb, Successive Rejects and Thompson sampling are evaluated and their
performance compared to existing approaches. The results show that multi-
armed bandit pruning methods, especially those based on UCB, outperform other
pruning methods.

• Journal Article Type: Article
• Acceptance Date: Jul 9, 2019
• Online Publication Date: Sep 26, 2019
• Publication Date: Jul 17, 2020
• Deposit Date: Jul 10, 2019
• Publicly Available Date: Oct 9, 2019
• Journal: The Computer Journal
• Print ISSN: 0010-4620
• Electronic ISSN: 1460-2067
• Publisher: Oxford University Press
• Volume: 63
• Issue: 7
• Pages: 1099-1108
• DOI: https://doi.org/10.1093/comjnl/bxz078
• Publisher URL: https://doi.org/10.1093/comjnl/bxz078
• Related Public URLs: https://academic.oup.com/comjnl