Visual Sensitivity Analysis for Artificial Neural Networks

  • Roberto Therón
  • Juan Francisco De Paz
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 4224)


A challenge in ANN research is how to reduce the number of inputs to the model in high dimensional problems, so it can be efficiently applied. The ANNs black-box operation makes not possible to explain the relationships between features and inputs. Some numerical methods, such as sensitivity analysis, try to fight this problem. In this paper, we combine a sensitivity analysis with a linked multi-dimensional visualization that takes advantage of user interaction, providing and efficient way to analyze and asses both the dimension reduction results and the ANN behavior.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Berrar, D.P., Downes, C.S., Dubitzky, W.: Multiclass cancer classification using gene expression profiling and probabilistic neural. In: Pacific Symposium on Biocomputing, vol. 8, pp. 5–16 (2003)Google Scholar
  2. 2.
    Sjoberg, J., Zhang, Q., Ljung, L., Benveniste, A., Delyon, B., Glorennec, P., Hjalmarsson, H., Juditsky, A.: Nonlinear black-box modeling in system identification a unified overview. Automatica 31, 1691–1724 (1995)CrossRefMathSciNetGoogle Scholar
  3. 3.
    Gevrey, M., Dimopoulos, I., Lek, S.: Review and comparison of methods to study the contribution of variables in artificial neural network models. Ecological Modelling 160, 249–264 (2003)CrossRefGoogle Scholar
  4. 4.
    Inselberg, A.: The plane with parallel coordinates. The Visual Computer 1, 69–91 (1985)zbMATHCrossRefGoogle Scholar
  5. 5.
    Tzeng, F.Y., Ma, K.L.: Opening the black box - data driven visualization of neural networks. In: Proceedings of IEEE Visualization 2005 Conference, pp. 383–390 (2005)Google Scholar
  6. 6.
    Craven, M., Shavlik, J.: Visualizing learning and computation an artificial neural networks. International Journal on Artificial Intelligence Tools 1, 399–425 (1992)CrossRefGoogle Scholar
  7. 7.
    Streeter, M.J., Ward, M.O., Alvarez, S.A.: Nvis: An interactive visualization tool for neural networks. In: Proceedings of SPIE Symposium on Visual Data Exploration and Analysis, pp. 234–241 (2001)Google Scholar
  8. 8.
    Duch, W.: Visualization of hidden node activity in neural networks: I and ii. In: Proceedings of the International Conference on Artificial Intelligence and Soft Computing, pp. 38–49 (2004)Google Scholar
  9. 9.
    Cook, R.D., Yin, X.: Special invited paper: Dimension reduction and visualization in discriminant analysis (with discussion). Australian & New Zealand Journal of Statistics 43, 147–199 (2001)zbMATHCrossRefMathSciNetGoogle Scholar
  10. 10.
    Hwang, J.N., Choi, J.J., Oh, S., II, R.J.M.: Query-based learning applied to partially trained multilayer perceptrons. IEEE Transactions on Neural Networks 2, 131–136 (1991)Google Scholar
  11. 11.
    Fu, L., Chen, T.: Sensitivity analysis for input vector in multilayer feedforward networks. In: Proceedings of IEEE International Conference on Neural Networks, vol. 1, pp. 215–218 (1993)Google Scholar
  12. 12.
    Keim, D.A.: Information visualization and visual data mining. IEEE Transactions on Visualization and Computer Graphics 8, 1–8 (2002)CrossRefGoogle Scholar
  13. 13.
    Becker, R.A., Cleveland, W.S.: Brushing scatterplots. Technometrics 29, 127–142 (1987)CrossRefMathSciNetGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2006

Authors and Affiliations

  • Roberto Therón
    • 1
  • Juan Francisco De Paz
    • 1
  1. 1.Departamento de Informática y AutomáticaFacultad de Ciencias – Universidad de SalamancaSalamancaSpain

Personalised recommendations