Abstract
Selective attention learning is proposed to improve the speed of the error backpropagation algorithm of a multilayer Perceptron. Class-selective relevance for evaluating the importance of a hidden node in an off-line stage and a node attention technique for measuring the local errors appearing at the output and hidden nodes in an on-line learning process are employed to selectively update the weights of the network. The acceleration of learning time is then achieved by lowering the computational cost required for learning. By combining this method with other types of improved learning algorithms, further improvement in learning speed is also achieved. The effectiveness of the proposed method is demonstrated by the speaker adaptation task of an isolated word recognition system. The experimental results show that the proposed selective attention technique can reduce the adaptation time more than 65% in an average sense.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
Jacobs, R.A.: Increased Rates of Convergence Through Learning Rate Adaptation, Neural Networks, 1 (1988) 295–307
Park, D.J., Jun, B.E., Kim, J.H.: Novel Fast Training Algorithm for Multilayer Feedforward Neural Network, Electronic Letters, 28(6) (1992) 543–544
Fahlman, S.E.: Faster-Learning Variations on Backpropagation: An Empirical Study, Proc. Connectionist Models Summer School, Carnegie Mellon University, (1988)38–51
Oh, S.H.: Improving the Error Back-Propagation Algorithm with a Modified Error Function, IEEE Trans. Neural Networks, 8(3) (1997) 799–803
Mozer, M.C., Smolensky, P.: Using Relevance to Reduce Network Size Automatically, Connection Science, 1(1) (1989) 3–16
Abrash, V., Franco, H., Sankar, A., Cohen, M.: Connectionist Speaker Normalization and Adaptation, Proc. European Conf. Speech Communication and Technology, (1995) 2183–2186
Lee, C.H., Lin, C.H., Juang, B.H.: A Study on Speaker Adaptation of the Parameters of Continuous Density Hidden Markov Models, IEEE Trans. Signal Processing, 39(4) (1991) 806–814
Erdogan, S.S., Ng, G.S., Patrick, K.H.C.: Measurement Criteria for Neural Network Pruning,” Proc. IEEE TENCON Digital Signal Processing Applications, 1 (1996) 83–89
Rumelhart, D.E., Hinton, G.E., Williams, R.J.: Learning Internal Representations by Error Propagation, in Parallel Distributed Processing, Cambridge, MA: MIT Press, (1986)318–364
Davis, S.B., Mermelstein, P.: Comparison of Parametric Representations for Monosyllabic Word Recognition in Continuously Spoken Sentences, IEEE Trans. Acoustics, Speech, and Signal Processing, 28(4) (1980) 357–366
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2002 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Kim, IC., Chien, SI. (2002). Computational Cost Reduction by Selective Attention for Fast Speaker Adaptation in Multilayer Perceptron. In: Hendtlass, T., Ali, M. (eds) Developments in Applied Artificial Intelligence. IEA/AIE 2002. Lecture Notes in Computer Science(), vol 2358. Springer, Berlin, Heidelberg. https://doi.org/10.1007/3-540-48035-8_3
Download citation
DOI: https://doi.org/10.1007/3-540-48035-8_3
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-43781-9
Online ISBN: 978-3-540-48035-8
eBook Packages: Springer Book Archive