Multi-task learning by hierarchical Dirichlet mixture model for sparse failure prediction

Abstract

Sparsity and noisy labels occur inherently in real-world data. Previously, strong assumptions were made by domain experts to use their experience and expertise to select parameters for their models. Similar approach has been adopted in machine learning for hyper-parameter setting. However, these assumptions are often subjective and are not necessarily the optimal choice. To address this problem, we propose a data-driven approach to automate model parameter learning via a Bayesian nonparametric formulation. We propose hierarchical Dirichlet process mixture model (HDPMM) as a multi-task learning framework. It is used to learn the common parameters across different datasets in the same industry. In our experiments, we verified the capability of HDPMM for multi-task learning in infrastructure failure predictions. It was done by combining HDPMM with hierarchical beta process, which is our failure prediction model. In particular, multi-task learning was used to gain additional knowledge from failure records of water supply networks managed by other utility companies to improve prediction accuracy of our model. Notably, we have achieved superior accuracy for sparse predictions than previous state-of-the-art models. Moreover, we have demonstrated the capability of our proposed model in supporting preventive maintenance of critical infrastructure.

This is a preview of subscription content, access via your institution.

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5

References

  1. 1.

    Bishop, C.M., et al.: Neural Networks for Pattern Recognition. Oxford University Press, Oxford (1995)

    Google Scholar 

  2. 2.

    Bonilla, E.V., Chai, K.M.A., Williams, C.K.: Multi-task gaussian process prediction. NIPs 20, 153–160 (2007)

    Google Scholar 

  3. 3.

    Dai, W., Yang, Q., Xue, G.R., Yu, Y.: Self-taught clustering. In: Proceedings of the 25th international conference on Machine Learning, pp. 200–207. ACM (2008)

  4. 4.

    David, C.R., et al.: Regression models and life tables (with discussion). J. R. Stat. Soc. 34, 187–220 (1972)

    Google Scholar 

  5. 5.

    Friedman, J.H.: Stochastic gradient boosting. Comput. Stat. Data Anal. 38(4), 367–378 (2002)

    MathSciNet  Article  Google Scholar 

  6. 6.

    Gupta, S., Phung, D., Venkatesh, S.: Factorial multi-task learning: a Bayesian nonparametric approach. In: International Conference on Machine Learning, pp. 657–665 (2013)

  7. 7.

    Hjort, N.L., et al.: Nonparametric bayes estimators based on beta processes in models for life history data. Ann. Stat. 18(3), 1259–1294 (1990)

    MathSciNet  Article  Google Scholar 

  8. 8.

    Huelsenbeck, J.P., Jain, S., Frost, S.W., Pond, S.L.K.: A dirichlet process model for detecting positive selection in protein-coding DNA sequences. Proc. Natl. Acad. Sci. 103(16), 6263–6268 (2006)

    Article  Google Scholar 

  9. 9.

    Ibrahim, J.G., Chen, M.H., Sinha, D.: Bayesian Survival Analysis. Wiley Online Library, New York (2005)

    Google Scholar 

  10. 10.

    Kabir, G., Tesfamariam, S., Sadiq, R.: Predicting water main failures using bayesian model averaging and survival modelling approach. Reliab. Eng. Syst. Saf. 142, 498–514 (2015)

    Article  Google Scholar 

  11. 11.

    Kemp, C., Tenenbaum, J.B., Griffiths, T.L., Yamada, T., Ueda, N.: Learning systems of concepts with an infinite relational model. In: AAAI, vol. 3, p. 5 (2006)

  12. 12.

    Kettler, A., Goulter, I.: An analysis of pipe breakage in urban water distribution networks. Can. J. Civ. Eng. 12(2), 286–293 (1985)

    Article  Google Scholar 

  13. 13.

    Kleiner, Y., Rajani, B.: Comprehensive review of structural deterioration of water mains: statistical models. Urban Water 3(3), 131–150 (2001)

    Article  Google Scholar 

  14. 14.

    Kumar, A., Rizvi, S.A.A., Brooks, B., Vanderveld, R.A., Wilson, K.H., Kenney, C., Edelstein, S., Finch, A., Maxwell, A., Zuckerbraun, J., et al.: Using machine learning to assess the risk of and prevent water main breaks. (2018). arXiv preprint arXiv:1805.03597

  15. 15.

    Le Gat, Y., Eisenbeis, P.: Using maintenance records to forecast failures in water networks. Urban Water 2(3), 173–181 (2000)

    Article  Google Scholar 

  16. 16.

    Li, B., Zhang, B., Li, Z., Wang, Y., Chen, F., Vitanage, D.: Prioritising water pipes for condition assessment with data analytics. Australia’s International Water Conference & Exhibition (OzWater) (2015)

  17. 17.

    Li, Z., Zhang, B., Wang, Y., Chen, F., Taib, R., Whiffin, V., Wang, Y.: Water pipe condition assessment: a hierarchical beta process approach for sparse incident data. Mach. Learn. 95(1), 11–26 (2014)

    MathSciNet  Article  Google Scholar 

  18. 18.

    Lin, P., Zhang, B., Wang, Y., Li, Z., Li, B., Wang, Y., Chen, F.: Data driven water pipe failure prediction: a Bayesian nonparametric approach. In: Proceedings of the 24th ACM International on Conference on Information and Knowledge Management, pp 193–202. ACM (2015)

  19. 19.

    Luo, S., Chu, V.W., Zhou, J., Chen, F., Wong, R.K., Huang, W.: A multivariate clustering approach for infrastructure failure predictions. In: 2017 IEEE International Congress on Big Data (BigData Congress), pp. 274–281. IEEE (2017)

  20. 20.

    Luo, S., Chu, V.W., Li, Z., Wang, Y., Zhou, J., Chen, F., Wong, R.K.: Multitask learning for sparse failure prediction. In: Pacific-Asia Conference on Knowledge Discovery and Data Mining, pp. 3–14. Springer, Berlin (2019)

    Google Scholar 

  21. 21.

    Mailhot, A., Pelletier, G., Noël, J.F., Villeneuve, J.P.: Modeling the evolution of the structural state of water pipe networks with brief recorded pipe break histories: methodology and application. Water Resources Res. 36(10), 3053–3062 (2000)

    Article  Google Scholar 

  22. 22.

    Mavin, K.: Predicting the Failure Performance of Individual Water Mains. Urban Water Research Association of Australia, Sydney (1996)

    Google Scholar 

  23. 23.

    Misiūnas, D.: Failure monitoring and asset condition assessment in water supply systems. Vilniaus Gedimino technikos universitetas, Vilnius (2008)

  24. 24.

    Morris Jr., R.: Principal causes and remedies of water main breaks. J. Am. Water Works Assoc. 59(7), 782–798 (1967)

    Article  Google Scholar 

  25. 25.

    Pan, S.J., Yang, Q.: A survey on transfer learning. IEEE Trans. Knowl. Data Eng. 22(10), 1345–1359 (2009)

    Article  Google Scholar 

  26. 26.

    Pelletier, G., Mailhot, A., Villeneuve, J.P.: Modeling water pipe breaks–three case studies. J. Water Resources Plan. Manag. 129(2), 115–123 (2003)

    Article  Google Scholar 

  27. 27.

    Pitman, J., Yor, M., et al.: The two-parameter poisson-dirichlet distribution derived from a stable subordinator. Ann. Probab. 25(2), 855–900 (1997)

    MathSciNet  Article  Google Scholar 

  28. 28.

    Schwaighofer, A., Tresp, V., Yu, K.: Learning Gaussian process kernels via hierarchical bayes. In: Advances in Neural Information Processing Systems 17 (NIPS 2004), pp. 1209–1216 (2005)

  29. 29.

    Sethuraman, J.: A constructive definition of Dirichlet priors. Stat. Sin. 4, 639–650 (1994)

  30. 30.

    Shamir, U., Howard, C., et al.: An analytical approach to scheduling pipe replacement. J. Am. Water Works Assoc. 71(5), 248–258 (1979)

    Article  Google Scholar 

  31. 31.

    Teh, Y.W., Jordan, M.I., Beal, M.J., Blei, D.M.: Hierarchical dirichlet processes. J. Am. Stat. Assoc. 101(476), 1566–1581 (2006)

    MathSciNet  Article  Google Scholar 

  32. 32.

    Thibaux, R., Jordan, M.I.: Hierarchical beta processes and the indian buffet process. AISTATS 2, 564–571 (2007)

    Google Scholar 

  33. 33.

    Xue, Y., Liao, X., Carin, L., Krishnapuram, B.: Multi-task learning for classification with dirichlet process priors. J. Mach. Learn. Res. 8(Jan), 35–63 (2007)

    MathSciNet  MATH  Google Scholar 

  34. 34.

    Zhang, Y., Yang, Q.: A survey on multi-task learning. arXiv preprint arXiv:1707.08114 (2017)

Download references

Author information

Affiliations

Authors

Corresponding author

Correspondence to Simon Luo.

Ethics declarations

Conflict of interest

On behalf of all authors, the corresponding author states that there is no conflict of interest.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Luo, S., Chu, V.W., Li, Z. et al. Multi-task learning by hierarchical Dirichlet mixture model for sparse failure prediction. Int J Data Sci Anal (2020). https://doi.org/10.1007/s41060-020-00219-z

Download citation

Keywords

  • Multi-task learning
  • Sparse predictions
  • Dirichlet process
  • Beta process
  • Failure predictions