Effective parallelisation for machine learning

Kamp, Michael; Boley, Mario; Missura, Olana; G�rtner, Thomas

Abstract

We present a novel parallelisation scheme that simplifies the adaptation of learning algorithms to growing amounts of data as well as growing needs for accurate and confident predictions in critical applications. In contrast to other parallelisation techniques, it can be applied to a broad class of learning algorithms without further mathematical derivations and without writing dedicated code, while at the same time maintaining theoretical performance guarantees. Moreover, our parallelisation scheme is able to reduce the runtime of many learning algorithms to polylogarithmic time on quasi-polynomially many processing units. This is a significant step towards a general answer to an open question [21] on efficient parallelisation of machine learning algorithms in the sense of Nick’s Class (NC). The cost of this parallelisation is in the form of a larger sample complexity. Our empirical study confirms the potential of our parallelisation scheme with fixed numbers of processors and instances in realistic application scenarios.

Citation

Kamp, M., Boley, M., Missura, O., & Gärtner, T. (2017). Effective parallelisation for machine learning.

• Conference Name: 31st Annual Conference: Neural Information Processing Systems 2017
• End Date: Dec 9, 2017
• Acceptance Date: Sep 4, 2017
• Publication Date: Dec 4, 2017
• Deposit Date: Nov 24, 2017
• Publicly Available Date: Dec 4, 2017
• Journal: Advances in Neural Information Processing Systems
• Electronic ISSN: 1049-5258
• Publisher: Massachusetts Institute of Technology Press
• Peer Reviewed: Peer Reviewed
• Volume: 30
• Public URL: https://nottingham-repository.worktribe.com/output/898587
• Publisher URL: https://papers.nips.cc/paper/7226-effective-parallelisation-for-machine-learning
• Related Public URLs: https://nips.cc/Conferences/2017/Schedule?showEvent=9417
• Additional Information: Acknowlegement of acceptance for publication.; ; Effective Parallelisation for Machine Learning. Advances in Neural Information Processing Systems 30 (NIPS 2017). 31st Annual Conference: Neural Information Processing Systems 2017 held 4-9 December 2017, Long Beach, California.