In this paper we show how the survival analysis problem can be formulated in terms of support vector regression, starting from a quantile regression perspective. We define an appropriate weighted loss function which takes into account possibly censored observations, and we prove bounds on the estimation error and on the quantile property. We deduce that censoring is a limiting factor in the accuracy of solutions, though the overall rate of convergence is O(n − 1/2). Finally we show some applications of the model to synthetic data, and to the German Breast Cancer Study Group 2 data.


Survival analysis censored data support vector machines statistical learning theory quadratic optimisation 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Cox, D.R., Oakes, D.: Analysis of Survival Data. Chapman and Hall, London (1984)Google Scholar
  2. 2.
    Bishop, C.M.: Neural networks for pattern recognition. Oxford University Press, Oxford (1995)Google Scholar
  3. 3.
    Eleuteri, A., Tagliaferri, R., Milano, L., De Placido, S., De Laurentiis, M.: A novel neural network-based survival analysis model. Neural Networks 16, 855–864 (2003)CrossRefGoogle Scholar
  4. 4.
    Biganzoli, E., Boracchi, P., Mariani, L., Marubini, E.: Feed forward neural networks for the analysis of censored survival data: a partial logistic regression approach. Statistics in Medicine 17, 1169–1186 (1998)CrossRefGoogle Scholar
  5. 5.
    Ripley, B.D., Ripley, R.M.: Neural Networks as Statistical Methods in Survival Analysis. In: Dybowsky, R., Gant, V. (eds.) Artificial Neural Networks: Prospects for Medicine. Landes Biosciences Publishers (1998)Google Scholar
  6. 6.
    Watanabe, S.: Algebraic Analysis for Non-identifiable Learning Machines. Neural Computation 13, 899–933 (2001)zbMATHCrossRefGoogle Scholar
  7. 7.
    Koenker, R.: Censored Quantile Regression Redux. Journal of Statistical Software 27 (2008)Google Scholar
  8. 8.
    Bottai, M., Zhang, J.: Laplace regression with censored data. Biometrical Journal 52, 487–503 (2010)MathSciNetzbMATHCrossRefGoogle Scholar
  9. 9.
    Vapnik, V.N.: Statistical Learning Theory. John Wiley & Sons, New York (1998)zbMATHGoogle Scholar
  10. 10.
    Vapnik, V.N.: The Nature of Statistical Learning Theory. Springer, Heidelberg (1995)zbMATHGoogle Scholar
  11. 11.
    Van Belle, V., Pelckmans, K., Suykens, J.A.K., Van Huffel, S.: Survival SVM: a Practical Scalable Algorithm. In: Proceedings of the 16th European Symposium on Artifical Neural Networks, ESANN 2008 (2008)Google Scholar
  12. 12.
    Shivaswamy, P.K., Chu, W., Jansche, M.: A Support Vector Approach to Censored Targets. In: Proceedings of the 7th International Conference on Data Mining (2007)Google Scholar
  13. 13.
    Harrell Jr., F.E.: Regression Modelling Strategies. Springer, Heidelberg (2005)Google Scholar
  14. 14.
    Van Belle, V., Pelckmans, K., Suykens, J.A.K., Van Huffel, S.: Additive survival least-squares support vector machines. Statistics in Medicine 29, 296–308 (2010)MathSciNetGoogle Scholar
  15. 15.
    Lama, N., Boracchi, P., Biganzoli, E.: Partial logistic relevance vector machines in survival analysis. Journal of Applied Statistics 38(11) (2011)Google Scholar
  16. 16.
    Williams, C.K.I., Rasmussen, C.E.: Gaussian Processes for Machine Learning. The MIT Press (2006)Google Scholar
  17. 17.
    Eleuteri, A., Taktak, A.F.G.: Survival time prediction by support vector regression. Proceedings of Royal Liverpool University Hospital Research Open Day (2007)Google Scholar
  18. 18.
    Eleuteri, A.: Support vector survival regression. In: 4th IET Conference on Advances in Medical, Signal and Information Processing (2008)Google Scholar
  19. 19.
    Koenker, R.: Quantile Regression. Cambridge University Press (2005)Google Scholar
  20. 20.
    Takeuchi, I., Le, Q.V., Sears, T.D., Smola, A.J.: Nonparametric Quantile Estimation. Journal of Machine Learning Research 7, 1231–1264 (2006)MathSciNetzbMATHGoogle Scholar
  21. 21.
    Carbonez, A., Györfi, G., van der Meulin, E.C.: Partition-estimate of a regression function under random censoring. Statist. Decis. 13, 21–27 (1995)zbMATHGoogle Scholar
  22. 22.
    Bartlett, P.L., Mendelson, S.: Rademacher and Gaussian complexities: Risk bounds and structural results. Journal of Machine Learning Research 3, 463–482 (2002)MathSciNetGoogle Scholar
  23. 23.
    Wang, J.L., Stute, W.: The strong law under random censorship. Annals of Statistics 21, 14–44 (1993)MathSciNetCrossRefGoogle Scholar
  24. 24.
    Bartlett, P.L., Bousquet, O., Mendelson, S.: Localized Rademacher Complexities. In: Kivinen, J., Sloan, R.H. (eds.) COLT 2002. LNCS (LNAI), vol. 2375, pp. 44–58. Springer, Heidelberg (2002)CrossRefGoogle Scholar
  25. 25.
    Christmann, A., Steinwart, I.: Consistency of kernel-based quantile regression. Journal of Applied Stochastic Models in Business and Industry 24, 171–183 (2008)MathSciNetzbMATHCrossRefGoogle Scholar
  26. 26.
    Chernozukov, V., Fernandez-Val, I., Galichon, A.: Improving point and interval estimators of monotone functions by rearrangement. Biometrika 96, 559–575 (2009)MathSciNetCrossRefGoogle Scholar
  27. 27.
    Schumacher, M., Basert, G., Bojar, H., Huebner, K., Olschewski, M., Sauerbrei, W., Schmoor, C., Beyerle, C., Neumann, R.L.A., Rauschecker, H.F.: For the German Breast Cancer Study Group: Randomized 2x2 trial evaluating hormonal treatment and the duration of chemotherapy in node-positive breast cancer patients. Journal of Clinical Oncology 12, 2086–2093 (1994)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2012

Authors and Affiliations

  • Antonio Eleuteri
    • 1
  • Azzam F. G. Taktak
    • 1
  1. 1.Department of Medical Physics and Clinical EngineeringRoyal Liverpool and Broadgreen University Hospitals NHS TrustLiverpoolUnited Kingdom

Personalised recommendations