Abstract
This chapter deals with the application of saliency analysis to Support. Vector Machines (SVMs) for feature selection. The importance of feature is ranked by evaluating the sensitivity of the network output to the feature input in terms of the partial derivative. A systematic approach to remove irrelevant features based on the sensitivity is developed. Two simulated non-linear time series and five real financial time series are examined in the experiment. The simulation results show that that saliency analysis is effective in SVMs for identifying important features.
This is a preview of subscription content, log in via an institution to check access.
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
Muller, R., Smola, J. A., Scholkopf, B. (1997): Prediction Time Series with Support Vector Machines. In: Proceedings of International Conference on Artificial Neural Networks, 999
Mukherjee, S., Osuna, E., Girosi, F. (1997): Nonlinear Prediction of Chaotic Time Series Using Support Vector Machines. Proc. Of IEEE NNSP’97, Amelia Isl, FL
Vapnik, V. N., Golowich, S. E., Smola, A. J. (1996): Support Vector Method for Function Approximation, Rregression Estimation and Signal Processing. Advances in Neural Information Processing Systems, Vol. 9, 281–287
Muller, K. R., Smola, J. A., Ratsch, G., Scholkopf, B., Kohlmorgen, J. (1999): Prediction Time Series with Support Vector Machines. Advances in Kernel Methods, The MIT Press
Barzilay, O., Brailovsky, V. L. (1999): On Domain Knowledge and Feature Selection Using a Support Vector Machine. Pattern Recognition Letters, Vol. 20, 475–484
Steppe, J. M., Bauer, Jr. K. W. (1997): Feature Saliency Measures. Computers Math. Application, Vol. 33, 109–126
Reed, R. (1993): Pruning Algorithms-a Survey. IEEE Transactions on Neural Networks, Vol. 4, 740–747
Ruck, D. W., Rogers, S. K., Kabrisky, M. (1990): Feature Selection Using a Multilayer Perceptron. Journal of Neural Network Computing, Vol. 2, 40–48
Belue, L. M., Bauer, Jr. K. W. (1995): Determining Input Features for Multi-layer Perceptrons. Neurocomputing, Vol. 7, 111–121
Stepp, J. M., Bauer, Jr. K. W. (1996): Improved Feature Screening in Feed-forward Neural Networks. Neurocomputing, 1347–58
Zurada, M. J., Malinowski, A., Usui, S. (1997): Perturbation Method for Deleting Redundant inputs of Perceptron Networks. Neurocomputing, Vol. 14, 177193
Vapnik, V. N. (1995): The Nature of Statistical Learning Theory. Springer-Verlag
Smola, A. J., Scholkopf, B. (1998): A Tutorial on Support Vector Regression. NeuroCOLT Technical Report TR, Royal Holloway College
Smola, A. J. (1998): Learning With Kernels. Ph.D. Thesis GMD, Birlinghoven
Cibas, T., Soulie, F. F., Gallinari, P., Raudys, S. (1996): Variable Selection with Neural Networks. Neurocomputing, Vol. 12, 223–248
Sexton, R. S. (1998): Identifying Irrelevant Input Variables in Chaotic Time Series Problems: Using a Genetic Algorithm for Training Neural Networks. Journal of Computational Intelligence in Finance, 34–41
Murray, W. H. (1992): Microsoft C/C++7: The Complete Reference. Osborne McGraw-Il ill
Thomason, M. (1999): The Practitioner Methods and Tool. Journal of Computational Intelligence in Finance, Vol. 7, No. 3, 36–45
Thomason, M. (1999): The Practitioner Methods and Tool. Journal of Computational Intelligence in Finance, Vol. 7, No. 4, 35–45
Thomason, M. (1999): The Practitioner Methods and Tool. Journal of Computational intelligence in Finance, Vol. 7, No. 6, 35–45
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2004 Springer-Verlag Berlin Heidelberg
About this chapter
Cite this chapter
Cao, L., Tay, F.E.H. (2004). Saliency Analysis of Support Vector Machines for Feature Selection in Financial Time Series Forecasting. In: Chen, SH., Wang, P.P. (eds) Computational Intelligence in Economics and Finance. Advanced Information Processing. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-662-06373-6_7
Download citation
DOI: https://doi.org/10.1007/978-3-662-06373-6_7
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-642-07902-3
Online ISBN: 978-3-662-06373-6
eBook Packages: Springer Book Archive