Abstract
A misclassification rate is most often used as a feature selection criterion. However, in the cases, when the numerical force of the training set is not sufficiently large in relation to the number of features, the risk of choosing the noisy features is very high. It produces difference between error estimations derived on the basis of the training and testing sets, so the error rate estimation can not be sufficiently confident. Feature preselection based on analysis of dependences between features is recommended in such types of tasks. An advantage of this approach is shown in the paper. As a feature selection criterion the Pearson chi-square statistics has been used.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
Similar content being viewed by others
References
Duda RO, Hart PE (2000) Pattern Classification (2nd ed.), Wiley
Chernoff H, Lehmann EL (1954) The use of maximum likelihood estimates in Χ2 tests for goodness-of-fit, The Annals of Mathematical Statistics
Rosner BA (2005) Fundamentals of Biostatistics, 6th ed., Duxbury Press
Tadeusiewicz R, Izworski A, Majewski J (1993) Biometria, Wyd. AGH
Jozwik A, Chmielewski L, Sklodowski M, Cudny W (1998) A parallel net of (1-NN, k-NN) classifier for optical inspection of surface defects in ferrites, Machine Graphics & Vision, vol. 7, no. 1–2, pp. 99–112
Newman DJ, Hettich S, Blake CL, Merz CJ (1998) UCI Repository of machine learning databases [http://www.ics.uci.edu/~mlearn/MLRepository.html], Irvine, CA: University of California, Department of Information and Computer Science
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2007 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Kosla, P. (2007). A Feature Selection Approach in Problems with a Great Number of Features. In: Kurzynski, M., Puchala, E., Wozniak, M., Zolnierek, A. (eds) Computer Recognition Systems 2. Advances in Soft Computing, vol 45. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-540-75175-5_50
Download citation
DOI: https://doi.org/10.1007/978-3-540-75175-5_50
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-75174-8
Online ISBN: 978-3-540-75175-5
eBook Packages: EngineeringEngineering (R0)