Abstract
Stochastic gradient descent is the most prevalent algorithm to train neural networks. However, other approaches such as evolutionary algorithms are also applicable to this task. Evolutionary algorithms bring unique trade-offs that are worth exploring, but computational demands have so far restricted exploration to small networks with few parameters. We implement an evolutionary algorithm that executes entirely on the GPU, which allows to efficiently batch-evaluate a whole population of networks. Within this framework, we explore the limited evaluation evolutionary algorithm for neural network training and find that its batch evaluation idea comes with a large accuracy trade-off. In further experiments, we explore crossover operators and find that unprincipled random uniform crossover performs extremely well. Finally, we train a network with 92k parameters on MNIST using an EA and achieve 97.6% test accuracy compared to 98% test accuracy on the same network trained with Adam. Code is available at https://github.com/jprellberg/gpuea.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
A tensor is a multi-dimensional array.
References
Baioletti, M., Di Bari, G., Poggioni, V., Tracolli, M.: Can differential evolution be an efficient engine to optimize neural networks? In: Nicosia, G., Pardalos, P., Giuffrida, G., Umeton, R. (eds.) MOD 2017. LNCS, vol. 10710, pp. 401–413. Springer, Cham (2018). https://doi.org/10.1007/978-3-319-72926-8_33
Beyer, H.: Evolutionary algorithms in noisy environments: theoretical issues and guidelines for practice. In: Computer Methods in Applied Mechanics and Engineering, pp. 239–267 (1998)
Das, S., Mullick, S.S., Suganthan, P.: Recent advances in differential evolution: an updated survey. Swarm Evol. Comput. 27(Complete), 1–30 (2016). https://doi.org/10.1016/j.swevo.2016.01.004
Desell, T.: Large scale evolution of convolutional neural networks using volunteer computing. In: Proceedings of the Genetic and Evolutionary Computation Conference Companion (GECCO 2017), pp. 127–128. ACM, New York (2017). https://doi.org/10.1145/3067695.3076002
Floreano, D., Dürr, P., Mattiussi, C.: Neuroevolution: from architectures to learning. Evol. Intell. 1(1), 47–62 (2008). https://doi.org/10.1007/s12065-007-0002-4
García-Pedrajas, N., Ortiz-Boyer, D., Hervás-Martínez, C.: An alternative approach for neural network evolution with a genetic algorithm: crossover by combinatorial optimization. Neural Netw. 19(4), 514–528 (2006). https://doi.org/10.1016/j.neunet.2005.08.014, http://www.sciencedirect.com/science/article/pii/S0893608005002297
Glorot, X., Bengio, Y.: Understanding the difficulty of training deep feedforward neural networks. In: Teh, Y.W., Titterington, M. (eds.) Proceedings of the Thirteenth International Conference on Artificial Intelligence and Statistics. Proceedings of Machine Learning Research. PMLR, vol. 9, pp. 249–256, Chia Laguna Resort, Sardinia, Italy, 13–15 May 2010. http://proceedings.mlr.press/v9/glorot10a.html
Ioffe, S., Szegedy, C.: Batch normalization: accelerating deep network training by reducing internal covariate shift. In: Proceedings of the 32nd International Conference on Machine Learning (ICML 2015), Lille, France, pp. 448–456 (2015)
Kingma, D.P., Ba, J.: Adam: a method for stochastic optimization. In: The International Conference on Learning Representations (ICLR 2015), December 2015
Liu, H., Simonyan, K., Vinyals, O., Fernando, C., Kavukcuoglu, K.: Hierarchical representations for efficient architecture search. In: International Conference on Learning Representations (ICML 2018) abs/1711.00436 (2018). http://arxiv.org/abs/1711.00436
Morse, G., Stanley, K.O.: Simple evolutionary optimization can rival stochastic gradient descent in neural networks. In: Proceedings of the Genetic and Evolutionary Computation Conference (GECCO 2016), pp. 477–484. ACM, New York (2016). https://doi.org/10.1145/2908812.2908916
Real, E., et al.: Large-scale evolution of image classifiers. In: Proceedings of the 34th International Conference on Machine Learning (ICML 2017) (2017). https://arxiv.org/abs/1703.01041
Stanley, K.O., D’Ambrosio, D.B., Gauci, J.: A hypercube-based encoding for evolving large-scale neural networks. Artif. Life 15(2), 185–212 (2009). https://doi.org/10.1162/artl.2009.15.2.15202
Thierens, D.: Non-redundant genetic coding of neural networks. In: Proceedings of IEEE International Conference on Evolutionary Computation, pp. 571–575, May 1996. https://doi.org/10.1109/ICEC.1996.542662
Yaman, A., Mocanu, D.C., Iacca, G., Fletcher, G., Pechenizkiy, M.: Limited evaluation cooperative co-evolutionary differential evolution for large-scale neuroevolution. In: Genetic and Evolutionary Computation Conference (GECCO 2018) (2018)
Zhang, X., Clune, J., Stanley, K.O.: On the relationship between the OpenAI evolution strategy and stochastic gradient descent. CoRR abs/1712.06564 (2017). http://arxiv.org/abs/1712.06564
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2018 Springer Nature Switzerland AG
About this paper
Cite this paper
Prellberg, J., Kramer, O. (2018). Limited Evaluation Evolutionary Optimization of Large Neural Networks. In: Trollmann, F., Turhan, AY. (eds) KI 2018: Advances in Artificial Intelligence. KI 2018. Lecture Notes in Computer Science(), vol 11117. Springer, Cham. https://doi.org/10.1007/978-3-030-00111-7_23
Download citation
DOI: https://doi.org/10.1007/978-3-030-00111-7_23
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-00110-0
Online ISBN: 978-3-030-00111-7
eBook Packages: Computer ScienceComputer Science (R0)