Asynchronous parallel hyperparameter search with population evolution

被引：0

作者：

Jiang Y.-L. ^{[1
,2
]}

Zhao K. ^{[3
]}

Cao J.-J. ^{[4
]}

Fan J. ^{[4
]}

Liu Y. ^{[4
]}

机构：

[1] School of Information Engineering, Huzhou University, Huzhou

[2] Zhejiang Province Key Laboratory of Smart Management & Application of Modern Agricultural Resources, Huzhou University, Huzhou

[3] College of Mathematics and Computer Science, Zhejiang Normal University, Jinhua

[4] Institute of Cyber Systems and Control, Zhejiang University, Hangzhou

来源：

Kongzhi yu Juece/Control and Decision | 2021年 / 36卷 / 08期

关键词：

Asynchronous parallelism; Deep learning; Evolutionary algorithm; Hyperparameter search; Parallel framework; Population;

D O I：

10.13195/j.kzyjc.2019.1743

中图分类号：

学科分类号：

摘要：

In recent years, with the continuous increase of deep learning models, especially deep reinforcement learning models, the training cost, that is, the search space of hyperparameters, has also continuously increased. However, most traditional hyperparameter search algorithms are based on sequential execution of training, which often takes weeks or even months to find a better hyperparameter configuration. In order to solve the problem of the long search time hyperparameters and the difficulty in finding a better hyperparameter of deep reinforcement learning configuration, this paper proposes a new hyper-parameter search algorithm, named asynchronous parallel hyperparameter search with population evolution. This algorithm combines the idea of evolutionary algorithms and uses a fixed resource budget to search the population model and its hyperparameters asynchronously and in parallel, thereby improving the performance of the algorithm. It is realized that a parameter search algorithm can run on the Ray parallel distributed framework. Experiments show that the parametric asynchronous parallel search based on population evolution on the parallel framework is better than the traditional hyperparameter search algorithm, and its performance is stable. Copyright ©2021 Control and Decision.

引用

页码：1825 / 1833

页数：8

共 26 条

[1] Wistuba M, Schilling N, Schmidt-Thieme L., Hyperparameter search space pruning — A new component for sequential model-based hyperparameter optimization, Proceedings of the 2015 European Conference on Machine Learning and Knowledge Discovery in Databases, pp. 104-119, (2015)
[2] Hinton G E, Osindero S, Teh Y W., A fast learning algorithm for deep belief nets, Neural Computation, 18, 7, pp. 1527-1554, (2006)
[3] Thakur N, Kumar S, Patle V K., Comparison of serial and parallel searching in multicore systems, Proceedings of 2015 International Conference on Parallel, Distributed and Grid Computing, pp. 11-13, (2015)
[4] Yang E S, Kim J D, Park C Y, Et al., Hyperparameter tuning for hidden unit conditional random fields, Engineering Computations, 34, 6, pp. 2054-2062, (2017)
[5] Rahnama A H A, Toloo M, Zaidenberg N J., An LP-based hyperparameter optimization model for language modeling, The Journal of Supercomputing, 74, 5, pp. 2151-2160, (2018)
[6] Jamieson K, Talwalkar A., Non-stochastic best arm identification and hyperparameter optimization, Proceedings of the 19th International Conference on Artificial Intelligence and Statistics, pp. 240-248, (2016)
[7] Li L, Jamieson K, Desalvo G, Et al., Hyperband: A novel bandit-based approach to hyperparameter optimization, Journal of Machine Learning Research, 18, pp. 1-52, (2018)
[8] Hutter F, Hoos H H, Leyton-Brown K., Sequential model-based optimization for general algorithm configuration, Proceedings of the 5th International Conference on Learning and Intelligent Optimization, pp. 507-523, (2011)
[9] Bergstra J, Bardenet R, Bengio Y, Et al., Algorithms for hyper-parameter optimization, Proceedings of the 24th International Conference on Neural Information Processing Systems, pp. 2546-2554, (2011)
[10] Li L, Jamieson K, Rostamizadeh A, Et al., Massively parallel hyperparameter tuning, (2018)

← 1 2 3 →