Abstract
Solution of inverse problems is usually sensitive to noise in the input data, as problems of this type are usually ill-posed or ill-conditioned. While neural networks have high noise resilience by themselves, this may be not enough in case of incorrect inverse problems. In their previous studies, the authors have demonstrated that the method of group determination of parameters, as well as noise addition during training of a neural network, can improve the resilience of the solution to noise in the input data. This study is devoted to the investigation of joint application of these methods. It has been performed on a model problem, for which the direct function is set explicitly as a polynomial.
This study has been performed at the expense of the grant of Russian Science Foundation (Project No. 14-11-00579).
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Zhdanov, M.: Inverse Theory and Applications in Geophysics, 2nd edn. Elsevier, Amsterdam (2015). 730 pp
Yagola, A., Kochikov, I., Kuramshina, G.: Inverse Problems of Vibrational Spectroscopy. De Gruyter, Berlin (1999). 297 pp
Orlov, V., Zinchenko, V., Ushakov, V., Velichkovsky, B.: Physiological noise reduction algorithms for fMRI data. Procedia Comput. Sci. 123, 334–340 (2018)
Isaev, I.V., Obornev, E.A., Obornev, I.E., Shimelevich, M.I., Dolenko, S.A.: Increase of the resistance to noise in data for neural network solution of the inverse problem of magnetotellurics with group determination of parameters. Lect. Notes Comput. 9886, 502–509 (2016)
Isaev, I.V., Vervald, E., Sarmanova, O., Dolenko, S.A.: Neural network solution of an inverse problem in Raman spectroscopy of multi-component solutions of inorganic salts: group determination as a method to increase noise resilience of the solution. Procedia Comput. Sci. 123, 177–182 (2018)
Isaev, I.V., Dolenko, S.A.: Training with noise as a method to increase noise resilience of neural network solution of inverse problems. Opt. Mem. Neural Netw. 25(3), 142–148 (2016)
Isaev, I.V., Dolenko, S.A.: Adding noise during training as a method to increase resilience of neural network solution of inverse problems: test on the data of magnetotelluric sounding problem. Stud. Comput. Intell. 736, 9–16 (2018)
Holmstrom, L., Koistinen, P.: Using additive noise in back-propagation training. IEEE Trans. Neural Netw. 3(1), 24–38 (1992)
Matsuoka, K.: Noise injection into inputs in back-propagation learning. IEEE Trans. Syst. Man Cybern. 22(3), 436–440 (1992)
Bishop, C.M.: Training with noise is equivalent to Tikhonov regularization. Neural Comput. 7(1), 108–116 (1995)
An, G.: The effects of adding noise during backpropagation training on a generalization performance. Neural Comput. 8(3), 643–674 (1996)
Zur, R.M., Jiang, Y., Pesce, L.L., Drukker, K.: Noise injection for training artificial neural networks: a comparison with weight decay and early stopping. Med. Phys. 36(10), 4810–4818 (2009)
Wang, C., Principe, J.C.: Training neural networks with additive noise in the desired signal. IEEE Trans. Neural Netw. 10(6), 1511–1517 (1999)
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
Isaev, I., Dolenko, S. (2019). Group Determination of Parameters and Training with Noise Addition: Joint Application to Improve the Resilience of the Neural Network Solution of a Model Inverse Problem to Noise in Data. In: Samsonovich, A. (eds) Biologically Inspired Cognitive Architectures 2018. BICA 2018. Advances in Intelligent Systems and Computing, vol 848. Springer, Cham. https://doi.org/10.1007/978-3-319-99316-4_18
Download citation
DOI: https://doi.org/10.1007/978-3-319-99316-4_18
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-99315-7
Online ISBN: 978-3-319-99316-4
eBook Packages: Intelligent Technologies and RoboticsIntelligent Technologies and Robotics (R0)