Scoring Systems and Large Margin Perceptron Ranking using Positive Weights

Bernd-Jürgen Falkowski, Arne-Michael Törsel



Large Margin Perceptron learning with positive coefficients is proposed in the context of so-called scoring systems used for assessing creditworthiness as stipulated in the Basel II central banks capital accord of the G10-states. Thus a potential consistency problem can be avoided. The approximate solution of a related ranking problem using a modified large margin algorithm producing positive weights is described. Some experimental results obtained from a Java prototype are exhibited. An important parallelization using Java concurrent programming is sketched. Thus it becomes apparent that combining the large margin algorithm presented here with the pocket algorithm can provide an attractive alternative to the use of support vector machines. Related algorithms are briefly discussed.


  1. Banking Committee on Banking Supervision: International Convergence of Capital Measurements and Capital Standards, A Revised Framework, Bank for International Settlements,, April 18th, 2006
  2. Bishop,C.M.: Neural Networks for Pattern Recognition. OUP, (1998)
  3. Block, H.D.; Levin, S.A.: On the Boundedness of an Iterative Procedure for Solving a System of Linear Inequalities. Proc. AMS, (1970)
  4. Cover, T.M.: Geometrical and Statistical Properties of Systems of Linear Inequalities with Applications in Pattern Recognition. IEEE Trans. on Electronic Computers, Vol. 14, (1965)
  5. Crammer, K.; Singer, Y.: Pranking with Ranking, NIPS, (2001)
  6. Episcopos, A.; Pericli, A.; Hu, J.: Commercial Mortgage Default: A Comparison of the Logistic Model with Artificial Neural Networks. Proceedings of the 3rd Internl .Conference on Neural Networks in the Capital Markets, London, England, (1995)
  7. Falkowski, B.-J.: Lernender Klassifikator, Offenlegungsschrift DE 101 14874 A1, Deutsches Patent- und Markenamt, München, (Learning Classifier, Patent Number DE 101 14874 A1, German Patent Office, Munich) (2002)
  8. Falkowski, B.-J.: On a Ranking Problem Associated with Basel II. Journal of Information and Knowledge Management, Vol. 5, No. 4, (2006), invited article.
  9. Falkowski, B.-J.: A Note on a Large Margin Perceptron Algorithm, Information Technology and Control, Vol. 35, No. 3 A, (2006)
  10. B.-J. Falkowski; M. Appelt; C. Finger; S. Koch; H. van der Linde: Scoring Systems and Large Margin Perceptron Ranking. In: Proceedings of the 18th IRMA Conference, Vol. 2, Ed. Mehdi Khosrow-Pour, IGI Publish. Hershey PA USA, (2007)
  11. Gallant, S.I.: Perceptron-based Learning Algorithms. IEEE Transactions on Neural Networks, Vol. I, No. 2, (1990)
  12. Hand, D.J.; Henley, W.E.: Statistical Classification Methods in Consumer Credit Scoring: a Review. Journal of the Royal Statistical Society, Series A, 160, Part 3, (1997)
  13. Herbrich, R.; Graepel, T.; Obermayer, K.: Large Margin Rank Boundaries for Ordinal Regression. In: Advances in Large Margin Classifiers (Eds. Smola, A.J.; Bartlett, P.; Schölkopf, B.; Schuurmans, D.), MIT Press, Neural Information Processing Series, (2000)
  14. Krauth, W.; Mezard, M.: Learning Algorithms with Optimal Stability in Neural Networks. J. Phys. A: Math. Gen. 20, (1987)
  15. Mathieson, M.: Ordinal Models for Neural Networks. Neural Networks in Financial Engineering, Proceedings of the 3rd International Conference on Neural Networks in the Capital Markets, World Scientific, (1996)
  16. Minsky, M.L.; Papert, S.: Perceptrons. MIT Press, (Expanded Edition 1990)
  17. Müller, M.; Härdle, W.: Exploring Credit Data. In: Bol, G.; Nakhneizadeh, G.; Racher, S.T.; Ridder, T.; Vollmer, K.-H. (Eds.): Credit Risk-Measurement, Evaluation, and Management, Physica-Verlag, (2003)
  18. Muselli, M.: On Convergence Properties of Pocket Algorithm. IEEE Trans. on Neural Networks, 8 (3), 1997
  19. Shadbolt, J.; Taylor, J.G.(Eds.): Neural Networks and the Financial Markets. SpringerVerlag, (2002)
  20. Shashua, A.; Levin, A.: Taxonomy of Large Margin Principle Algorithms for Ordinal Regression Problems. Tech. Report 2002-39, Leibniz Center for Research, School of Comp. Science and Eng., the Hebrew University of Jerusalem, (2002)
  21. Shashua, A.; Levin, A.:Ranking with Large Margin Principle: Two Approaches, NIPS 14, (2003)
  22. Shawe-Taylor, J.; Cristianini, N.: Kernel Methods for Pattern Analysis. Cambridge University Press, (2004)
  23. Thomas, L.C.: A Survey of Credit and Behavioural Scoring: Forecasting Financial Risk of Lending to Consumers. Internl. Journal of Forecasting, 16, (2000)
  24. Vapnik, V.N.: Statistical Learning Theory. John Wiley & Sons, (1998)
  25. Wong, S.K.M.; Ziarko, W.; Wong, P.C.N.: Generalized Vector Space Model in Information Retrieval. Proceedings of the 8th ACM SIGIR Conference on Research and Development in Information Retrieval, USA, (1985)
  26. Xu, J.; Zhang, X.; Li, Y.: Large Margin Kernel Pocket Algorithm. In: Proceedings of the International Joint Conference on Neural Networks 2001, IJCNN'01, Vol. 2. New York: IEEE, (2001)

Paper Citation

in Harvard Style

Falkowski B. and Törsel A. (2008). Scoring Systems and Large Margin Perceptron Ranking using Positive Weights . In Proceedings of the 8th International Workshop on Pattern Recognition in Information Systems - Volume 1: PRIS, (ICEIS 2008) ISBN 978-989-8111-42-5, pages 213-222. DOI: 10.5220/0001733702130222

in Bibtex Style

author={Bernd-Jürgen Falkowski and Arne-Michael Törsel},
title={Scoring Systems and Large Margin Perceptron Ranking using Positive Weights},
booktitle={Proceedings of the 8th International Workshop on Pattern Recognition in Information Systems - Volume 1: PRIS, (ICEIS 2008)},

in EndNote Style

JO - Proceedings of the 8th International Workshop on Pattern Recognition in Information Systems - Volume 1: PRIS, (ICEIS 2008)
TI - Scoring Systems and Large Margin Perceptron Ranking using Positive Weights
SN - 978-989-8111-42-5
AU - Falkowski B.
AU - Törsel A.
PY - 2008
SP - 213
EP - 222
DO - 10.5220/0001733702130222