Skip to main content
Disciplines
Automotive Business IT + Informatics Construction + Real Estate Electrical Engineering + Electronics Energy + Sustainability Insurance + Risk Finance + Banking Management + Leadership Marketing + Sales Mechanical Engineering + Materials
Events
DE
EN
Books
Journals
Topic Page
Marketing
Start single access now
Access for companies
EXTENDED SEARCH
Log in
Top
Published in:
Effective Self Attention Modeling for Aspect Based Sentiment Analysis Read chapter Read first chapter

2019 | OriginalPaper | Chapter

Parallel Algorithm Based on Singular Value Decomposition for High Performance Training of Neural Networks

Authors : Gabriele Maria Lozito, Valentina Lucaferri, Mauro Parodi, Martina Radicioni, Francesco Riganti Fulginei, Alessandro Salvini

Published in: Computational Science – ICCS 2019

Publisher: Springer International Publishing

Log in

Activate our intelligent search to find suitable subject content or patents.

search-config
loading …

Abstract

Neural Networks (NNs) are frequently applied to Multi Input Multi Output (MIMO) problems, where the amount of data to manage is extremely high and, hence, the computational time required for the training process is too large. Therefore, MIMO problems are often split into Multi Input Single Output (MISO) problems; MISOs are further decomposed into several Single Input Single Output (SISO) problems. The aim of this paper is to present an optimized approach for NNs training based on properties of Singular Value Decomposition (SVD), allowing to decompose the MISO NN into a collection of SISO NNs. The decomposition provides a two-fold advantage: firstly, each SISO NN can be trained by using a one-dimensional function, namely a limited dataset, and then a parallel architecture can be implemented on a PC-cluster, decreasing the computational cost. The parallel algorithm performance are validated by using magnetic hysteresis dataset with the aim to prove the computational speed up by preserving the accuracy.

Dont have a licence yet? Then find out more about our products and how to get one now:

Springer Professional "Wirtschaft+Technik"

Online-Abonnement

Mit Springer Professional "Wirtschaft+Technik" erhalten Sie Zugriff auf:

  • über 102.000 Bücher
  • über 537 Zeitschriften

aus folgenden Fachgebieten:

  • Automobil + Motoren
  • Bauwesen + Immobilien
  • Business IT + Informatik
  • Elektrotechnik + Elektronik
  • Energie + Nachhaltigkeit
  • Finance + Banking
  • Management + Führung
  • Marketing + Vertrieb
  • Maschinenbau + Werkstoffe
  • Versicherung + Risiko

Jetzt Wissensvorsprung sichern!

inform now

Springer Professional "Technik"

Online-Abonnement

Mit Springer Professional "Technik" erhalten Sie Zugriff auf:

  • über 67.000 Bücher
  • über 390 Zeitschriften

aus folgenden Fachgebieten:

  • Automobil + Motoren
  • Bauwesen + Immobilien
  • Business IT + Informatik
  • Elektrotechnik + Elektronik
  • Energie + Nachhaltigkeit
  • Maschinenbau + Werkstoffe




 

Jetzt Wissensvorsprung sichern!

inform now

Springer Professional "Wirtschaft"

Online-Abonnement

Mit Springer Professional "Wirtschaft" erhalten Sie Zugriff auf:

  • über 67.000 Bücher
  • über 340 Zeitschriften

aus folgenden Fachgebieten:

  • Bauwesen + Immobilien
  • Business IT + Informatik
  • Finance + Banking
  • Management + Führung
  • Marketing + Vertrieb
  • Versicherung + Risiko




Jetzt Wissensvorsprung sichern!

inform now
previous chapter Automating the Generation of Comparison Weights for Enhancing the AHP Decision-Making Process
next chapter In-Situ Visualization with Membrane Layer for Movie-Based Visualization
Literature
1.
Bizzarri, F., Parodi, M., Storace, M.: SVD-based approximations of bivariate functions. In: IEEE International Symposium on Circuits and Systems, vol. 5, p. 4915. IEEE 1999 (2005)
2.
Cardelli, E., Faba, A., Laudani, A., Riganti Fulginei, F., Salvini, A.: A neural approach for the numerical modeling of two-dimensional magnetic hysteresis. J. Appl. Phys. 117(17), 17D129 (2015)CrossRef
3.
Chen, M., Ge, S.S., How, B.V.E.: Robust adaptive neural network control for a class of uncertain mimo nonlinear systems with input nonlinearities. IEEE Trans. Neural Netw. 21(5), 796–812 (2010)CrossRef
4.
Duan, N., Xu, W., Wang, S., Zhu, J., Guo, Y.: Hysteresis modeling of high-temperature superconductor using simplified Preisach model. IEEE Trans. Magn. 51(3), 1–4 (2015)CrossRef
5.
Handgruber, P., Stermecki, A., Biro, O., Goričan, V., Dlala, E., Ofner, G.: Anisotropic generalization of vector Preisach hysteresis models for nonoriented steels. IEEE Trans. Magn. 51(3), 1–4 (2015)CrossRef
6.
Hovakimyan, N., Calise, A.J., Kim, N.: Adaptive output feedback control of a class of multi-input multi-output systems using neural networks. Int. J. Control 77(15), 1318–1329 (2004)MathSciNetCrossRef
7.
Huynh, H.T., Won, Y.: Training single hidden layer feedforward neural networks by singular value decomposition. In: Fourth International Conference on 2009 Computer Sciences and Convergence Information Technology, ICCIT 2009, pp. 1300–1304. IEEE (2009)
8.
Jianye, L., Yongchun, L., Jianpeng, B., Xiaoyun, S., Aihua, L.: Flaw identification based on layered multi-subnet neural networks. In: 2009 Second International Conference on Intelligent Networks and Intelligent Systems, pp. 118–121. IEEE (2009)
9.
Kabir, H., Wang, Y., Yu, M., Zhang, Q.J.: High-dimensional neural-network technique and applications to microwave filter modeling. IEEE Trans. Microw. Theory Tech. 58(1), 145–156 (2010)CrossRef
10.
Laudani, A., Lozito, G.M., Riganti Fulginei, F.: Dynamic hysteresis modelling of magnetic materials by using a neural network approach. In: 2014 AEIT Annual Conference-From Research to Industry: The Need for a More Effective Technology Transfer (AEIT), pp. 1–6. IEEE (2014)
11.
Laudani, A., Lozito, G.M., Riganti Fulginei, F., Salvini, A.: Modeling dynamic hysteresis through fully connected cascade neural networks. In: 2016 IEEE 2nd International Forum on Research and Technologies for Society and Industry Leveraging a better tomorrow (RTSI), pp. 1–5. IEEE (2016)
12.
Laudani, A., Salvini, A., Parodi, M., Riganti Fulginei, F.: Automatic and parallel optimized learning for neural networks performing mimo applications. Adv. Electr. Comput. Eng. 13(1), 3–12 (2013)CrossRef
13.
Makaveev, D., Dupré, L., De Wulf, M., Melkebeek, J.: Modeling of quasistatic magnetic hysteresis with feed-forward neural networks. J. Appl. Phys. 89(11), 6737–6739 (2001)CrossRef
14.
Rasilo, P., et al.: Modeling of hysteresis losses in ferromagnetic laminations under mechanical stress. IEEE Trans. Magn. 52(3), 1–4 (2016)CrossRef
15.
Riganti Fulginei, F., Salvini, A.: Neural network approach for modelling hysteretic magnetic materials under distorted excitations. IEEE Trans. Magn. 48(2), 307–310 (2012)CrossRef
16.
Riganti Fulginei, F., Salvini, A., Parodi, M.: Learning optimization of neural networks used for MIMO applications based on multivariate functions decomposition. Inverse Prob. Sci. Eng. 20(1), 29–39 (2012)MathSciNetCrossRef
17.
Salvini, A., Riganti Fulginei, F.: Genetic algorithms and neural networks generalizing the Jiles-Atherton model of static hysteresis for dynamic loops. IEEE Trans. Magn. 38(2), 873–876 (2002)CrossRef
18.
Salvini, A., Riganti Fulginei, F., Pucacco, G.: Generalization of the static Preisach model for dynamic hysteresis by a genetic approach. IEEE Trans. Magn. 39(3), 1353–1356 (2003)CrossRef
19.
Turk, C., Aradag, S., Kakac, S.: Experimental analysis of a mixed-plate gasketed plate heat exchanger and artificial neural net estimations of the performance as an alternative to classical correlations. Int. J. Therm. Sci. 109, 263–269 (2016)CrossRef
Metadata
Title
Parallel Algorithm Based on Singular Value Decomposition for High Performance Training of Neural Networks
Authors
Gabriele Maria Lozito
Valentina Lucaferri
Mauro Parodi
Martina Radicioni
Francesco Riganti Fulginei
Alessandro Salvini
Copyright Year
2019
Book
Computational Science – ICCS 2019

Print ISBN: 978-3-030-22749-4

Electronic ISBN: 978-3-030-22750-0

Copyright Year: 2019

DOI
https://doi.org/10.1007/978-3-030-22750-0_54

Premium Partner

    Image Credits