A Rejection Option for the Multilayer Perceptron Using Hyperplanes

  • Eduardo Gasca A.
  • Sergio Saldaña T.
  • José S. Sánchez G.
  • Valentín Velásquez G.
  • Eréndira Rendón L.
  • Itzel M. Abundez B.
  • Rosa M. Valdovinos R.
  • Rafael Cruz R.
Part of the Lecture Notes in Computer Science book series (LNCS, volume 6593)

Abstract

Currently, a growing quantity of the Artificial Intelligence tasks demand a high efficiency of the classification systems (classifiers); making an error in the classification of an object or event can cause serious problems. This is worrying when the classifiers confront tasks where the classes are not linearly separable, the classifiers efficiency diminishes considerably. One solution for decreasing this complication is the Rejection Option. In several circumstances it is advantageous to not have a decision be taken and wait to obtain additional information instead of making an error.

This work contains the description of a novel reject procedure whose purpose is to identify elements with a high risk of being misclassified; like those in an overlap zone. For this, the location of the object in evaluation is calculated with regard to two hyperplanes that emulate the classifiers decision boundary. The area between these hyperplanes is named an overlap region. If the element is localized in this area, it is rejected.

Experiments conducted with the artificial neural network Multilayer Perceptron, trained with the Backpropagation algorithm, show between 12.0%- 91.4%of the objects in question would have been misclassified if they had not been rejected.

Keywords

Reject option Multilayer Perceptron Backpropagation hyperplane overlap 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Burden, R.L., Douglas Faires, J., Reynolds, A.C.: Numerical Analysis. In: Prindle, Weber, Schmidt (eds.) Wadsworth International (1981)Google Scholar
  2. 2.
    Cordella, L.P., Stefano, C.D., Tortorella, F., Vento, M.: A Method for Improving Classification Reliability of Multilayer Perceptrons. IEEE Trans. on neural Networks 6(5), 1140–1147 (1995)CrossRefGoogle Scholar
  3. 3.
    Irie, B., Miyake, S.: Capabilities of three-layered perceptrons. In: Proc. of the IEEE Conference on Neural Networks, vol. I, pp. 641–648 (1988)Google Scholar
  4. 4.
    Lippmann, R.P.: An introduction to computing with neural nets. IEEE Acoust. Speech Signal Process. Mag. 4, 4–22 (1987)Google Scholar
  5. 5.
    Lin, C.-C., El-Jaroudi, A.: An algorithm to determine the feasibilities and weights of two-layer perceptrons for partitioning and classification. Pattern Recognition 31(11), 1613–1625 (1998)CrossRefGoogle Scholar
  6. 6.
    Looney, C.G.: Pattern Recognition Using Neural Networks: Theory and Algorithms for Engineers and Scientists. Oxford University Press, Oxford (1997)Google Scholar
  7. 7.
    Makhoul, J., El-Jaroudi, A., Schwartz, R.: Partitioning capabilities of two-layer neural networks. IEEE Trans. Signal Process. 39(6), 436–1440 (1991)CrossRefGoogle Scholar
  8. 8.
    Merz, C.J., Murphy, P.M.: Repository of Machine Learning Databases, University of California at Irvine (1998), http://www.csi.uci.edu/mlearn
  9. 9.
    Reister, D.B.: The least Squares Fit of a Hyperplane to Uncertain Data. ORNL (1996)Google Scholar
  10. 10.
    Shah, J.V., Poon, C.-S.: Linear independence of internal representations in multilayer perceptrons. IEEE Trans. on Neural Networks 10(1), 10–18 (1999)CrossRefGoogle Scholar
  11. 11.
    Singh, S., Markou, M.: An Approach to Novelty Detection Applied to the Classification of Images Regions. IEEE Trans. on Knowledge and Data Engineering 16(4), 396–407 (2004)CrossRefGoogle Scholar
  12. 12.
    De Stefano, C., Sansone, C., Vento, M.: To Reject or Not to Reject: That is the Question an Answer in Case of Neural Classifiers. IEEE Transactions on Systems, Man, and Cybernetics Part C Applications end Reviews 30(1), 84–94 (2000)CrossRefGoogle Scholar
  13. 13.
    Tsujitani, M., Koshimizu, T.: Neural Discriminant Analysis. IEEE Transactions on Neural Networks 11(6), 1394–1400 (2000)CrossRefGoogle Scholar
  14. 14.
    Wilson, C.L., Blue, J.L., Omidvar, O.M.: Training dynamics and neural network performance. Neural Networks 10(5), 907–923 (1997)CrossRefGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2011

Authors and Affiliations

  • Eduardo Gasca A.
    • 1
  • Sergio Saldaña T.
    • 1
  • José S. Sánchez G.
    • 2
  • Valentín Velásquez G.
    • 1
  • Eréndira Rendón L.
    • 1
  • Itzel M. Abundez B.
    • 1
  • Rosa M. Valdovinos R.
    • 2
  • Rafael Cruz R.
    • 1
  1. 1.Technological Institute of TolucaMexico
  2. 2.University Jaume ICastelloSpain

Personalised recommendations