Abstract
We consider the problem of locating small openings inside the domain of definition of elliptic equation using as the observation data the values of finite number of integral functionals. Application of neural networks requires a great number of training sets. The approximation of these functionals by means of topological derivative allows to generate training data very quickly. The results of computations for 2D examples show, that the method allows to determine an approximation of the global solution to the inverse problem, sufficiently closed to the exact solution.
Supported by the grant 4-T11A-015-24 of the State Committee for the Scientific Research of the Republic of Poland.
An erratum to this chapter can be found at http://dx.doi.org/10.1007/11550907_163 .
This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsPreview
Unable to display preview. Download preview PDF.
References
Barron, A.R.: Universal approximation bounds for superpositions of a sigmoidal function. IEEE Transactions on Information Theory 39, 930–945 (1993)
Hagan, M., Menhaj, M.: Training feedforward networks with the Marquardt algorithm. IEEE Trans. on Neural Networks 5, 989–993 (1994)
Il’in, A.M.: Matching of Asymptotic Expansions of Solutions of Boundary Value Problems. In: Translations of Mathematical Monographs, vol. 102. AMS, Providence (1992)
Jackowska, L., Sokołowski, J., Żochowski, A., Henrot, A.: On numerical solutions of shape inverse problems. Computational Optimization and Applications 23, 231–255 (2002)
Lewiński, T., Sokołowski, J., Żochowski, A.: Justification of the bubble method for the compliance minimization problems of plates and spherical shells CD-Rom. In: 3rd World Congress of Structural and Multidisciplinary Optimization (WCSMO-3) Buffalo/Niagara Falls, New York, May 17-21 (1999)
Nazarov, S.A., Sokołowski, J.: Asymptotic analysis of shape functionals, Les prépublications de l’Institut Élie Cartan, 51 (2001)
Roche, J.R., Sokołowski, J.: Numerical methods for shape identification problems. Special issue of Control and Cybernetics: Shape Optimization and Scientific Computations 5, 867–894 (1996)
Schiffer, M., Szegö, G.: Virtual mass and polarization. Transactions of the American Mathematical Society 67, 130–205 (1949)
Shumacher, A.: Topologieoptimierung von Bauteilstrukturen unter Verwendung von Lochpositionierungkriterien, Ph.D. Thesis, Universität–Gesamthochschule–Siegen, Siegen (1995)
Sokołowski, J., Żochowski, A.: On topological derivative in shape optimization. SIAM Journal on Control and Optimization 37(4), 1251–1272 (1999)
Sokołowski, J., Żochowski, A.: Topological derivative for optimal control problems. Control and Cybernetics 28(3), 611–626 (1999)
Sokołowski, J., Żochowski, A.: Topological derivatives for elliptic problems. Inverse Problems 15(1), 123–134 (1999)
Sokołowski, J., Żochowski, A.: Topological derivatives of shape functionals for elasticity systems. Mechanics of Structures and Machines 29, 333–351 (2001)
Sokołowski, J., Żochowski, A.: Optimality conditions for simultaneous topology and shape optimization. SIAM Journal on Control and Optimization 42(4), 1198–1221 (2003)
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2005 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Jackowska-Strumiłło, L., Sokołowski, J., Żochowski, A. (2005). Topological Derivative and Training Neural Networks for Inverse Problems. In: Duch, W., Kacprzyk, J., Oja, E., Zadrożny, S. (eds) Artificial Neural Networks: Formal Models and Their Applications – ICANN 2005. ICANN 2005. Lecture Notes in Computer Science, vol 3697. Springer, Berlin, Heidelberg. https://doi.org/10.1007/11550907_62
Download citation
DOI: https://doi.org/10.1007/11550907_62
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-28755-1
Online ISBN: 978-3-540-28756-8
eBook Packages: Computer ScienceComputer Science (R0)