Advertisement

Interpretability in Machine Learning – Principles and Practice

  • P. J. G. Lisboa
Part of the Lecture Notes in Computer Science book series (LNCS, volume 8256)

Abstract

Theoretical advances in machine learning have been reflected in many research implementations including in safety-critical domains such as medicine. However this has not been reflected in a large number of practical applications used by domain experts. This bottleneck is in a significant part due to lack of interpretability of the non-linear models derived from data. This lecture will review five broad categories of interpretability in machine learning - nomograms, rule induction, fuzzy logic, graphical models & topographic mapping. Links between the different approaches will be made around the common theme of designing interpretability into the structure of machine learning models, then using the armoury of advanced analytical methods to achieve generic non-linear approximation capabilities.

Keywords

Fuzzy Logic Latent Variable Model Rule Induction Machine Learning Model Predictive Inference 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Lisboa, P.J.G.: Industrial use of safety-related artificial neural networks. HSE CR 237/2001, HMSO (2001), http://www.hse.gov.uk/research/crr_pdf/2001/crr01327.pdf
  2. 2.
    Lisboa, P.J.G.: A review of evidence of health benefit from artificial neural networks in medical intervention. Neural Networks 15(1), 9–37 (2002)CrossRefGoogle Scholar
  3. 3.
    Lisboa, P.J.G., Taktak, A.F.G.: The use of artificial neural networks in decision support in cancer: A systematic review. Neural Networks 19(4), 408–415 (2006)CrossRefzbMATHGoogle Scholar
  4. 4.
    Chiu, S.: Developing commercial applications of intelligent control. IEEE Control Syst. Mag. 17(2), 94–100 (1997)MathSciNetCrossRefGoogle Scholar
  5. 5.
    Vellido, A., Lisboa, P.J.G.: Handling outliers in brain tumour MRS data analysis through robust topographic mapping. Computers in Biology and Medicine 36(10), 1049–1063 (2006)CrossRefGoogle Scholar
  6. 6.
    Van Belle, V., Lisboa, P.J.G.: Research Directions in Interpretable Machine Learning. In: European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning (ESANN), Bruges, April 24-26, pp. 191–196 (2013)Google Scholar
  7. 7.
    Breiman, L.: Statistical Modeling: The Two Cultures. Statistical Science 16(3), 199–231 (2001)MathSciNetCrossRefzbMATHGoogle Scholar
  8. 8.
    Bacciu, D., Lisboa, P.J.G., Sperdutti, A., Villmann, T.: Probabilistic Modelling in Machine Learning. In: Alippi, C., et al. (eds.) Handbook on Computational Intelligence. Springer (accepted, 2013)Google Scholar
  9. 9.
    Lisboa, P.J.G., Ellis, I.O., Green, A.R., Ambrogi, F., Dias, M.B.: Cluster-based visualisation with scatter matrices. Pattern Recognition Letters 29(13), 1814–1823 (2008)CrossRefGoogle Scholar
  10. 10.
    Bartholomew, Knott, Moustaki: Latent Variable Models and Factor Analysis: A Unified Approach, 3rd edn. (2011)Google Scholar
  11. 11.
    Gorban, A.N., Zinovyev, A.: Principal manifolds and graphs in practice: from molecular biology to dynamical systems. International Journal of Neural Systems 20(3), 219–232 (2010)CrossRefGoogle Scholar
  12. 12.
    Etchells, T.A., Lisboa, P.J.G.: Orthogonal search-based rule extraction (OSRE) from trained neural networks: A practical and efficient approach. IEEE Transactions on Neural Networks 17(2), 374–384 (2006)CrossRefGoogle Scholar
  13. 13.
    Rögnvaldsson, T., Etchells, T.A., You, L., Garwicz, D., Jarman, I.H., Lisboa, P.J.G.: How to find simple and accurate rules for viral protease cleavage specificities. BMC Bioinformatics 10, 149 (2009)CrossRefGoogle Scholar
  14. 14.
    Lisboa, P.J.G., Etchells, T.A., Pountney, D.C.: Minimal MLPs do not model the XOR logic. Neurocomputing, Rapid Communication 48(1-4), 1033–1037 (2002)CrossRefzbMATHGoogle Scholar
  15. 15.
    Jarman, I.H., Etchells, T.A., Martín, J.D., Lisboa, P.J.G.: An integrated framework for risk profiling of breast cancer patients following surgery. Artificial Intelligence in Medicine 42, 165–188 (2008)CrossRefGoogle Scholar
  16. 16.
    Bacciu, D., Etchells, T.A., Lisboa, P.J.G., Whittaker, J.: Efficient identification of independence networks using mutual information. Computational Statistics 28(2), 621–646 (2013)MathSciNetCrossRefGoogle Scholar
  17. 17.
    Fernandez, F., Duarte, A., Sanchez, A.: Optimization of the Fuzzy Partition of a Zero-order Takagi-Sugeno Model. In: Proc. Eleventh International Conference on Information Processing and Management of Uncertainty in Knowledge-based Systems (IPMU 2006), vol. I, pp. 898–905. Editions EDK (2006)Google Scholar
  18. 18.
    López de Mántaras, R., McSherry, D., Bridge, D., Leake, D., Smyth, B., Craw, S., Faltings, B., Maher, M.-L., Cox, M., Forbus, K., Keane, M., Aamodt, A., Watson, I.: Retrieval, reuse, revision, and retention in case-based reasoning. Knowledge Engineering Review 20(3), 215–240 (2005)CrossRefGoogle Scholar
  19. 19.
    Dutta, S., Bonissone, P.: Integrating Case Based And Rule Based Reasoning: The Possibilistic Connection. In: Proc. 6th Conference on Uncertainty in AI, Cambridge, MA, July 27-29, pp. 290–300 (1990)Google Scholar
  20. 20.
    Van Belle, V., Lisboa, P.J.G.: Automated Selection of Interaction Effects in Sparse Kernel Methods to Predict Pregnancy Viability. In: IEEE Symposium Series on Computational Intelligence, Singapore, April 16-19 (2013)Google Scholar
  21. 21.
    Ruiz, H., Etchells, T.A., Jarman, I.H., Martín, J.D., Lisboa, P.J.G.: A principled approach to network-based classification and data representation. Neurocomputing 112, 79–91 (2013)CrossRefGoogle Scholar
  22. 22.
    Newman, M.E.J.: Modularity and community structure in networks. Proc. Natl. Acad. Sci. USA 103(23), 8577–8582 (2006)CrossRefGoogle Scholar

Copyright information

© Springer International Publishing Switzerland 2013

Authors and Affiliations

  • P. J. G. Lisboa
    • 1
  1. 1.School of Computing and Mathematical SciencesLiverpool John Moores UniversityLiverpoolUK

Personalised recommendations