A fast neural network learning algorithm with approximate singular value decomposition

• Autorzy: Jankowski Norbert , Linowiecki Rafał

Identyfikatory

DOI

10.2478/amcs-2019-0043

• Języki publikacji: EN

Abstrakty

EN

The learning of neural networks is becoming more and more important. Researchers have constructed dozens of learning algorithms, but it is still necessary to develop faster, more flexible, or more accurate learning algorithms. With fast learning we can examine more learning scenarios for a given problem, especially in the case of meta-learning. In this article we focus on the construction of a much faster learning algorithm and its modifications, especially for nonlinear versions of neural networks. The main idea of this algorithm lies in the usage of fast approximation of the Moore–Penrose pseudo-inverse matrix. The complexity of the original singular value decomposition algorithm is O(mn²). We consider algorithms with a complexity of O(mnl), where l < n and l is often significantly smaller than n. Such learning algorithms can be applied to the learning of radial basis function networks, extreme learning machines or deep ELMs, principal component analysis or even missing data imputation.

Słowa kluczowe

EN

Moore–Penrose pseudoinverse; radial basis function network; extreme learning machine; kernel method; machine learning; singular value decomposition; deep extreme learning; principal component analysis

PL

pseudoodwrotność Moore–Penrose; radialna funkcja bazowa; maszyna uczenia ekstremalnego; uczenie maszynowe; analiza składników głównych

• Wydawca: Oficyna Wydawnicza Uniwersytetu Zielonogórskiego
• Czasopismo: International Journal of Applied Mathematics and Computer Science
• Rocznik: 2019
• Tom: Vol. 29, no. 3
• Strony: 581--594
• Opis fizyczny: Bibliogr. 21 poz., tab., wykr.

Twórcy

autor

Jankowski Norbert

• Department of Informatics, Faculty of Physics, Astronomy and Informatics, Nicolaus Copernicus University, ul. Grudziądzka 5, 87-100 Toruń, Poland

autor

Linowiecki Rafał

• Department of Informatics, Faculty of Physics, Astronomy and Informatics, Nicolaus Copernicus University, ul. Grudziądzka 5, 87-100 Toruń, Poland

Bibliografia

• [1] Bishop, C.M. (1991). Training with noise is equivalent to Tikhonov regularization, Neural Computation 7(1): 108–116.
• [2] Boser, B.E., Guyon, I.M. and Vapnik, V. (1992). A training algorithm for optimal margin classifiers, in D. Haussler (Ed.), Proceedings of the 5th Annual ACM Workshop on Computational Learning Theory, Pittsburgh, PA, USA, pp. 144–152.
• [3] Broomhead, D.S. and Lowe, D. (1988). Multivariable functional interpolation and adaptive networks, Complex Systems 2(3): 321–355.
• [4] Dumais, S.T. (2005). Latent semantic analysis, Annual Review of Information Science and Technology 38(1): 188–230.
• [5] Eirola, E., Lendasse, A., Vandewalle, V. and Biernacki, C. (2014). Mixture of Gaussians for distance estimation with missing data, Neurocomputing 131: 32–42.
• [6] Goodfellow, I., Bengio, Y. and Courville, A. (2016). Deep Learning, MIT Press, Cambridge, MA, http://www.deeplearningbook.org.
• [7] Górecki, T. and Łuczak, M. (2013). Linear discriminant analysis with a generalization of the Moore–Penrose pseudoinverse, International Journal of Applied Mathematics and Computer Science 23(2): 463–471, DOI: 10.2478/amcs-2013-0035.
• [8] Halko, N., Martinsson, P.G. and Tropp, J.A. (2011). Finding structure with randomness: Probabilistic algorithms for constructing approximate matrix decompositions, SIAM Review 53(2): 217–288.
• [9] Heseltine, T., Pears, N., Austin, J. and Chen, Z. (2003). Face recognition: A comparison of appearance-based approaches, 7th International Conference on Digital Image Computing: Techniques and Applications, Sydney, Australia, Vol. 1, pp. 59–68.
• [10] Huang, G.-B., Bai, Z., Kasun, L.L.C. and Vong, C.M. (2015). Local receptive fields based extreme learning machine, IEEE Computational Intelligence Magazine 10(2): 18–29.
• [11] Huang, G.-B., Zhu, Q.-Y. and Siew, C.-K. (2004). Extreme learning machine: A new learning scheme of feedforward neural networks, International Joint Conference on Neural Networks, Budapest, Hungary, pp. 985–990.
• [12] Huang, G.-B., Zhu, Q.-Y. and Siew, C.-K. (2006). Extreme learning machine: Theory and applications, Neurocomputing 70(1–3): 489–501.
• [13] Jankowski, N. (2013). Meta-learning and new ways in model construction for classification problems, Journal of Network & Information Security 4(4): 275–284.
• [14] Jankowski, N. (2018). Comparison of prototype selection algorithms used in construction of neural networks learned by SVD, International Journal of Applied Mathematics and Computer Science 28(4): 719–733, DOI: 10.2478/amcs-2018-0055.
• [15] Merz, C.J. and Murphy, P.M. (1998). UCI Repository of Machine Learning Databases, https://archive.ics.uci.edu/ml/index.php.
• [16] Mitchell, T. (1997). Machine Learning, McGraw Hill, New York, NY.
• [17] Rumelhart, D.E., Hinton, G.E. and Williams, R.J. (1986). Learning internal representations by error propagation, in J.L.M.D.E. Rumelhart (Ed.), Parallel Distributed Processing: Explorations in Microstructure of Congnition, Vol. 1: Foundations, MIT Press, Cambridge, MA, pp. 318–362.
• [18] Sovilj, D., Eirola, E., Miche, Y., Bjork, K.-M., Nian, R., Akusok, A. and Lendasse, A. (2016). Extreme learning machine for missing data using multiple imputations, Neurocomputing 174(PA): 220–231.
• [19] Tang, J., Deng, C., Member, S. and Huang, G.-B. (2016). Extreme learning machine for multilayer perceptron, IEEE Transactions on Neural Networks and Learning Systems 27(4): 809–821.
• [20] Tikhonov, A.N. and Arsenin, V.Y. (1977). Solutions of Ill-posed Problems, W.H.Winston, Washington, DC.
• [21] Vapnik, V. (1995). The Nature of Statistical Learning Theory, Springer-Verlag, New York, NY.

Uwagi

PL

Opracowanie rekordu w ramach umowy 509/P-DUN/2018 ze środków MNiSW przeznaczonych na działalność upowszechniającą naukę (2019).