Abstract
The problem of selection the most informative features is reduced to an optimization problem for the average risk functional whose maximization is equivalent to maximization of informational distance between distributions of features in two classes. We consider a maximization procedure for the average risk functional via empirical risk, estimating the divergence between them, with Rademacher complexity. The proposed method has been applied efficiently to problems of selection parameters important to separate the states of technological processes. We show an experimental comparison of the developed approach with other widely known feature selection techniques.
Similar content being viewed by others
References
Iverson, D.L., Data Mining Applications for Space Mission Operations System Health Monitoring, Proc. SpaceOps 2008 Conf., ESA, EUMETSAT, AIAA, Heidelberg, Germany, May 2008.
Kostyukov, V.N. and Naumenko, A.P., Analysis of Modern Methods and Means for Monitoring and Diagnostics of Pneumatic Pumps. Part 1. Online Monitoring Systems, V Mire NK, 2010, no. 1 (47), pp. 64–70.
Vapnik, V.N. and Chervonenkis, A.Ya., Teoriya raspoznavaniya obrazov (Image Recognition Theory), Moscow: Nauka, 1974.
Wolf, L. and Shashua, A., Features Selection for Unsupervised and Supervised Inference: The Emergence of Sparsity in a Weight-Based Approach, J. Machine Learning Res., 2005, vol. 6, pp. 1855–1887.
Blum, A. and Langley, P., Selection of Relevant Features and Examples in Machine Learning, AI, 1997, vol. 97, pp. 245–271.
John, G.H., Kohavi, R., and Pfleger, K., Irrelevant Features and the Subset Selection Problem, Proc. 11th Int. Conf. on Machine Learning, Morgan Kaufmann Publishers, 1994, pp. 121–129.
Kira, K. and Rendell, L., The Feature Selection Problem: Traditional Methods and a New Algorithm, 10th National Conf. on Artificial Intelligence, Cambridge: MIT Press, 1992, pp. 129–134.
Allmuallim, H. and Dietterich, T.G., Learning with Many Irrelevant Features, Proc. 9th National Conf. on Artificial Intelligence, San Jose: AAAI, 1991, pp. 547–552.
Jolliffe, I.T., Principal Component Analysis, New York: Springer-Verlag, 1986.
Comon, P., Independent Component Analysis. A New Concept, Signal Process., 1994, vol. 36, pp. 287–314.
Koller, D. and Sahami, M., Toward Optimal Feature Selection, Proc. 13th Int. Conf. on Machine Learning, Morgan Kaufmann Publishers, 1996, pp. 284–292.
Kullback, S. and Leibler, R.A., On Information and Sufficiency, Annals Math. Statist., 1951, vol. 22, no. 1, pp. 79–86.
Novovicova, J., Pudil, P., and Kittler, J., Divergence Based Feature Selection for Multimodal Class Densities, IEEE Trans. Patt. Anal. Machine Intelligen., 1996, vol. 18 (2), pp. 218–223.
Coetzee, F.M., Correcting Kullback–Leibler Distance for Feature Selection, Patt. Recognit. Lett., 2005, vol. 26, no. 11, pp. 1675–1683.
Eguchi, S. and Copas, J., Interpreting Kullback–Leibler Divergence with the Neyman–Pearson Lemma, J. Multivariate Anal., 2006, vol. 97, pp. 2034–2040.
Koltchinskii, V. and Panchenko, D., Rademacher Process and Bounding the Risk of Function Learning, in High Dimension. Probab. II, Gine, D.E., Wellner, J., Eds., Basel: Birkhauser, 1999, pp. 443–457.
Hall, M.A., Correlation-based Feature Selection for Discrete and Numeric Machine Learning, Proc. 17th Int. Conf. on Machine Learning (ICML-00), Morgan Kaufmann Publish., 2000.
IBM SPSS Modeler 14.2 Algorithms Guide, available at ftp://public.dhe.ibm.com/software/analytics/spss/documentation/modeler/.
Asuncion, A. and Newman, D.J., UCI Machine Learning Repository (http://www.ics.uci.edu/~mlearn/MLRepository.html), Irvine, CA: University of California, School of Information and Computer Science, 2007.
Tsurko, V.V. and Mikhal’skii, A.I., Statistical Analysis of the Relation between Cancer and Accompanying Diseases, Usp. Gerontologii, 2013, vol. 26, no. 4, pp. 766–774.
Author information
Authors and Affiliations
Corresponding author
Additional information
Original Russian Text © V.V. Tsurko, A.I. Michalski, 2016, published in Avtomatika i Telemekhanika, 2016, No. 12, pp. 136–154.
This paper was recommended for publication by L.A. Mironovskii, a member of the Editorial Board
Rights and permissions
About this article
Cite this article
Tsurko, V.V., Michalski, A.I. The contrast features selection with empirical data. Autom Remote Control 77, 2212–2226 (2016). https://doi.org/10.1134/S0005117916120109
Received:
Published:
Issue Date:
DOI: https://doi.org/10.1134/S0005117916120109