Abstract
Common types of artificial neural networks have been well known to suffer from the presence of outlying measurements (outliers) in the data. However, there are only a few available robust alternatives for training common form of neural networks. In this work, we investigate robust fitting of multilayer perceptrons, i.e. alternative approaches to the most common type of feedforward neural networks. Particularly, we consider robust neural networks based on the robust loss function of the least trimmed squares, for which we express formulas for derivatives of the loss functions. Some formulas, which are however incorrect, have been already available. Further, we consider a very recently proposed multilayer perceptron based on the loss function of the least weighted squares, which appears a promising highly robust approach. We also derive the derivatives of the loss functions, which are to the best of our knowledge a novel contribution of this paper. The derivatives may find applications in implementations of the robust neural networks, if a (gradient-based) backpropagation algorithm is used.
The research was financially supported by the projects GA19-05704S (J. Kalina) and GA18-23827S (P. Vidnerová) of the Czech Science Foundation.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Aladag, C.H., Egrioglu, E., Yolcu, U.: Robust multilayer neural network based on median neuron model. Neural Comput. Appl. 24, 945–956 (2014)
Beliakov, G., Kelarev, A., Yearwood, J.: Derivative-free optimization and neural networks for robust regression. Optimization 61, 1467–1490 (2012)
Cisse, M., Bojanowski, P., Grave, E., Dauphin, Y., Usunier, N.: Parseval networks: improving robustness to adversarial examples. In: Proceedings of the 34th International Conference on Machine Learning ICML 2017, pp. 854–863 (2017)
D’Urso, P., Leski, J.M.: Fuzzy \(c\)-ordered medoids clustering for interval-valued data. Pattern Recogn. 58, 49–67 (2016)
Frank, A., Asuncion, A.: UCI Machine Learning Repository. University of California, Irvine (2010). http://archive.ics.uci.edu/ml/
Haykin, S.O.: Neural Networks and Learning Machines: A Comprehensive Foundation, 2nd edn. Prentice Hall, Upper Saddle River (2009)
Henderson, D.J., Parmeter, C.F.: Applied Nonparametric Econometrics. Cambridge University Press, New York (2015)
Jurečková, J., Picek, J., Schindler, M.: Robust Statistical Methods with R, 2nd edn. Chapman & Hall/CRC, Boca Raton (2019)
Kalina, J.: A robust pre-processing of BeadChip microarray images. Biocybern. Biomed. Eng. 38, 556–563 (2018)
Kalina, J., Schlenker, A.: A robust supervised variable selection for noisy high-dimensional data. BioMed Res. Int. 2015, 1–10 (2015). Article no. 320385
Kalina, J., Vidnerová, P.: Robust training of radial basis function neural networks. In: Proceedings 18th International Conference ICAISC 2019. Lecture Notes in Computer Science, vol. 11508, pp. 113–124 (2019)
Kalina, J., Vidnerová, P.: On robust training of regression neural networks. In: Proceedings 5th International Workshop on Functional and Operatorial Statistics IWFOS (2020, accepted)
Kordos, M., Rusiecki, A.: Reducing noise impact on MLP training-techniques and algorithms to provide noise-robustness in MLP network training. Soft. Comput. 20, 46–65 (2016)
Paulheim, H., Meusel, R.: A decomposition of the outlier detection problem into a set of supervised learning problems. Mach. Learn. 100, 509–531 (2015)
Racine, J., Su, L., Ullah, A.: The Oxford Handbook of Applied Nonparametric and Semiparametric Econometrics and Statistics. Oxford University Press, Oxford (2014)
Rousseeuw, P.J., Van Driessen, K.: Computing LTS regression for large data sets. Data Min. Knowl. Disc. 12, 29–45 (2006)
Rusiecki, A.: Robust learning algorithm based on LTA estimator. Neurocomputing 120, 624–632 (2013)
Rusiecki, A., Kordos, M., Kamiński, T., Greń, K.: Training neural networks on noisy data. Lect. Notes Artif. Intell. 8467, 131–142 (2014)
Víšek, J.Á.: The least trimmed squares. Part I: consistency. Kybernetika 42, 1–36 (2006)
Víšek, J.Á.: Consistency of the least weighted squares under heteroscedasticity. Kybernetika 47, 179–206 (2011)
Acknowledgments
The authors would like to thank David Coufal and Jiří Tumpach for discussion.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2020 The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Kalina, J., Vidnerová, P. (2020). Robust Multilayer Perceptrons: Robust Loss Functions and Their Derivatives. In: Iliadis, L., Angelov, P., Jayne, C., Pimenidis, E. (eds) Proceedings of the 21st EANN (Engineering Applications of Neural Networks) 2020 Conference. EANN 2020. Proceedings of the International Neural Networks Society, vol 2. Springer, Cham. https://doi.org/10.1007/978-3-030-48791-1_43
Download citation
DOI: https://doi.org/10.1007/978-3-030-48791-1_43
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-48790-4
Online ISBN: 978-3-030-48791-1
eBook Packages: Computer ScienceComputer Science (R0)