Advertisement

Flash: A GP-GPU Ensemble Learning System for Handling Large Datasets

  • Ignacio Arnaldo
  • Kalyan Veeramachaneni
  • Una-May O’Reilly
Part of the Lecture Notes in Computer Science book series (LNCS, volume 8599)

Abstract

The Flash system runs ensemble-based Genetic Programming (GP) symbolic regression on a shared memory desktop. To significantly reduce the high time cost of the extensive model predictions required by symbolic regression, its fitness evaluations are tasked to the desktop’s GPU. Successive GP “instances” are run on different data subsets and randomly chosen objective functions. Best models are collected after a fixed number of generations and then fused with an adaptive, output-space method. New instance launches are halted once learning is complete. We demonstrate that Flash’s ensemble strategy not only makes GP more robust, but it also provides an informed online means of halting the learning process. Flash enables GP to learn from a dataset composed of 370K exemplars and 90 features, evolving a population of 1000 individuals over 100 generations in as few as 50 seconds.

Keywords

Genetic Programming GPGPU computing Ensembles 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Banzhaf, W., Harding, S., Langdon, W., Wilson, G.: Accelerating genetic programming through graphics processing units. In: Genetic Programming Theory and Practice VI. Genetic and Evolutionary Computation, pp. 1–19. Springer US (2009)Google Scholar
  2. 2.
    Bertin-Mahieux, T., Ellis, D.P., Whitman, B., Lamere, P.: The million song dataset. In: Proceedings of the 12th International Conference on Music Information Retrieval, ISMIR 2011 (2011)Google Scholar
  3. 3.
    Chitty, D.M.: A data parallel approach to genetic programming using programmable graphics hardware. In: Proceedings of the 9th Annual GECCO Conference, GECCO 2007, pp. 1566–1573. ACM, New York (2007)Google Scholar
  4. 4.
    Dijkstra, E.W.: Algol 60 translation. Supplement, Algol 60 Bulletin 10 (1960)Google Scholar
  5. 5.
    Harding, S., Banzhaf, W.: Fast genetic programming on GPUs. In: Ebner, M., O’Neill, M., Ekárt, A., Vanneschi, L., Esparcia-Alcázar, A.I. (eds.) EuroGP 2007. LNCS, vol. 4445, pp. 90–101. Springer, Heidelberg (2007)CrossRefGoogle Scholar
  6. 6.
    Harding, S., Banzhaf, W.: Implementing cartesian genetic programming classifiers on graphics processing units using GPU.NET. In: Proceedings of the 13th GECCO Conference, GECCO 2011, pp. 463–470. ACM, New York (2011)Google Scholar
  7. 7.
    Harding, S.L., Banzhaf, W.: Distributed genetic programming on GPUs using CUDA. In: Hidalgo, I., Fernandez, F., Lanchares, J. (eds.) PABA Workshop, Raleigh, NC, USA, September 13, pp. 1–10 (2009)Google Scholar
  8. 8.
    Kotanchek, M., Smits, G., Vladislavleva, E.: Trustable symbolic regression models: using ensembles, interval arithmetic and pareto fronts to develop robust and trust-aware models. In: Riolo, R., Soule, T., Worzel, B. (eds.) Genetic Programming Theory and Practice V. Genetic and Evolutionary Computation Series, pp. 201–220. Springer US (2008)Google Scholar
  9. 9.
    Langdon, W.B., Banzhaf, W.: A SIMD interpreter for genetic programming on GPU graphics cards. In: O’Neill, M., Vanneschi, L., Gustafson, S., Esparcia Alcázar, A.I., De Falco, I., Della Cioppa, A., Tarantino, E. (eds.) EuroGP 2008. LNCS, vol. 4971, pp. 73–85. Springer, Heidelberg (2008)CrossRefGoogle Scholar
  10. 10.
    Langdon, W.: A CUDA SIMT interpreter for genetic programming. Tech. Rep. TR-09-05, Department of Computer Science, Strand (June 2009) (revised)Google Scholar
  11. 11.
    Langdon, W.B.: A many threaded CUDA interpreter for genetic programming. In: Esparcia-Alcázar, A.I., Ekárt, A., Silva, S., Dignum, S., Uyar, A.Ş. (eds.) EuroGP 2010. LNCS, vol. 6021, pp. 146–158. Springer, Heidelberg (2010)CrossRefGoogle Scholar
  12. 12.
    Lewis, T.E., Magoulas, G.D.: Strategies to minimise the total run time of cyclic graph based genetic programming with GPUs. In: Proceedings of the 11th GECCO Conference, GECCO 2009, pp. 1379–1386. ACM, New York (2009)Google Scholar
  13. 13.
    Maitre, O., Querry, S., Lachiche, N., Collet, P.: EASEA parallelization of tree-based Genetic Programming. In: 2010 IEEE Congress on Evolutionary Computation (CEC), pp. 1–8 (2010)Google Scholar
  14. 14.
    Maitre, O., Lachiche, N., Collet, P.: Fast evaluation of GP trees on GPGPU by optimizing hardware scheduling. In: Esparcia-Alcázar, A.I., Ekárt, A., Silva, S., Dignum, S., Uyar, A.Ş. (eds.) EuroGP 2010. LNCS, vol. 6021, pp. 301–312. Springer, Heidelberg (2010)CrossRefGoogle Scholar
  15. 15.
    NVIDIA Corporation: NVIDIA CUDA C programming guide, version 3.2 (2010)Google Scholar
  16. 16.
    Robilliard, D., Marion-Poty, V., Fonlupt, C.: Population parallel GP on the G80 GPU. In: O’Neill, M., Vanneschi, L., Gustafson, S., Esparcia Alcázar, A.I., De Falco, I., Della Cioppa, A., Tarantino, E. (eds.) EuroGP 2008. LNCS, vol. 4971, pp. 98–109. Springer, Heidelberg (2008)CrossRefGoogle Scholar
  17. 17.
    Robilliard, D., Marion-Poty, V., Fonlupt, C.: Genetic programming on graphics processing units. Genetic Programming and Evolvable Machines 10(4), 447–471 (2009)CrossRefGoogle Scholar
  18. 18.
    Veeramachaneni, K., Derby, O., Sherry, D., O’Reilly, U.M.: Learning regression ensembles with genetic programming at scale. In: Proceeding of the Fifteenth GECCO Conference, GECCO 2013, pp. 1117–1124. ACM, New York (2013)Google Scholar
  19. 19.
    Wilson, G., Banzhaf, W.: Linear genetic programming GPGPU on Microsoft Xbox 360. In: IEEE Congress on Evolutionary Computation, pp. 378–385 (2008)Google Scholar
  20. 20.
    Yang, Y.: Adaptive regression by mixing. Journal of the American Statistical Association 96(454), 574–588 (2001)CrossRefzbMATHMathSciNetGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2014

Authors and Affiliations

  • Ignacio Arnaldo
    • 1
  • Kalyan Veeramachaneni
    • 1
  • Una-May O’Reilly
    • 1
  1. 1.Computer Science and Artificial Intelligence LaboratoryMITCambridgeUSA

Personalised recommendations