Abstract
The aim of this paper is to introduce two widely applicable regularization methods based on the direct modification of weight matrices. The first method, Weight Reinitialization, utilizes a simplified Bayesian assumption with partially resetting a sparse subset of the parameters. The second one, Weight Shuffling, introduces an entropy- and weight distribution-invariant non-white noise to the parameters. The latter can also be interpreted as an ensemble approach. The proposed methods are evaluated on benchmark datasets, such as MNIST, CIFAR-10 or the JSB Chorales database, and also on time series modeling tasks. We report gains both regarding performance and entropy of the analyzed networks. We also made our code available as a GitHub repository (https://github.com/rpatrik96/lod-wmm-2019).
The research presented in this paper has been supported by the European Union, co-financed by the European Social Fund (EFOP-3.6.2-16-2017-00013, Thematic Fundamental Research Collaborations Grounding Innovation in Informatics and Infocommunications), by the BME-Artificial Intelligence FIKP grant of Ministry of Human Resources (BME FIKP-MI/SC), by Doctoral Research Scholarship of Ministry of Human Resources (ÚNKP-19-4-BME-189) in the scope of New National Excellence Program, by János Bolyai Research Scholarship of the Hungarian Academy of Sciences. We gratefully acknowledge the support of NVIDIA Corporation with the donation of the Titan Xp GPU used for this research. Patrik Reizinger expresses his gratitude for the financial support of the Nokia Bell Labs Hungary.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Ignite (2019). https://github.com/pytorch/ignite
Bergstra, J., Bardenet, R., Bengio, Y., Kégl, B.: Algorithms for hyper-parameter optimization. In: Advances in Neural Information Processing Systems (NIPS), pp. 2546–2554 (2011). https://doi.org/2012arXiv1206.2944S
Cho, K., et al.: Learning Phrase Representations Using RNN Encoder-Decoder for Statistical Machine Translation. arXiv (2014). https://doi.org/10.1074/jbc.M608066200
Glorot, X., Bengio, Y.: Understanding the difficulty of training deep feedforward neural networks. PMLR 9, 249–256 (2010). https://doi.org/10.1.1.207.2059
Guennec, A.L., et al.: Data augmentation for time series classification using convolutional neural networks. In: ECML/PKDD Workshop on Advanced Analytics and Learning on Temporal Data (2016)
Hochreiter, S., Schmidhuber, J.: Long short-term memory. Neural Comput. 9(8), 1735–1780 (1997). https://doi.org/10.1162/neco.1997.9.8.1735
Ioffe, S., Szegedy, C.: Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift. arXiv (2015)
Krizhevsky, A., Hinton, G.: Learning multiple layers of features from tiny images. Technical report, Citeseer (2009). https://doi.org/10.1.1.222.9220
Krueger, D., et al.: Zoneout: Regularizing RNNs by Randomly Preserving Hidden Activations. arXiv, pp. 1–11 (2016). https://doi.org/10.1227/01.NEU.0000210260.55124.A4. http://arxiv.org/abs/1606.01305
LeCun, Y., Bottou, L., Bengio, Y., Haffner, P.: Gradient-based learning applied to document recognition. Proc. IEEE 86(11), 2278–2323 (1998). https://doi.org/10.1109/5.726791
Masters, D., Luschi, C.: Revisiting Small Batch Training for Deep Neural Networks. arXiv, pp. 1–18 (2018). https://doi.org/10.1016/j.biortech.2007.04.007. http://arxiv.org/abs/1804.07612
Neil, D., Pfeiffer, M., Liu, S.C.: Phased LSTM: accelerating recurrent network training for long or event-based sequences. In: Lee, D.D., Sugiyama, M., Luxburg, U.V., Guyon, I., Garnett, R. (eds.) Advances in Neural Information Processing Systems 29, pp. 3882–3890. Curran Associates, Inc. (2016)
Paszke, A., et al.: Automatic differentiation in PyTorch. In: Advances in Neural Information Processing Systems 30 (NIPS), pp. 1–4 (2017)
Plappert, M., et al.: Parameter Space Noise for Exploration. arXiv, pp. 1–18 (2017). http://arxiv.org/abs/1706.01905
Raś, Z., Wieczorkowska, A.: Advances in Music Information Retrieval. Studies in Computational Intelligence, vol. 274. Springer, Heidelberg (2010). https://doi.org/10.1007/978-3-642-11674-2
Srivastava, N., Hinton, G., Krizhevsky, A., Sutskever, I., Salakhutdinov, R.: Dropout: a simple way to prevent neural networks from overfitting. J. Mach. Learn. Res. 15, 1929–1958 (2014). https://doi.org/10.1214/12-AOS1000
Wan, L., Zeiler, M., Zhang, S., LeCun, Y., Fergus, R.: Regularization of neural networks using DropConnect. In: ICML, pp. 109–111 (2013). https://doi.org/10.1109/TPAMI.2017.2703082
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2019 Springer Nature Switzerland AG
About this paper
Cite this paper
Reizinger, P., Gyires-Tóth, B. (2019). Stochastic Weight Matrix-Based Regularization Methods for Deep Neural Networks. In: Nicosia, G., Pardalos, P., Umeton, R., Giuffrida, G., Sciacca, V. (eds) Machine Learning, Optimization, and Data Science. LOD 2019. Lecture Notes in Computer Science(), vol 11943. Springer, Cham. https://doi.org/10.1007/978-3-030-37599-7_5
Download citation
DOI: https://doi.org/10.1007/978-3-030-37599-7_5
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-37598-0
Online ISBN: 978-3-030-37599-7
eBook Packages: Computer ScienceComputer Science (R0)