ESNemble: an Echo State Network-based ensemble for workload prediction and resource allocation of Web applications in the cloud

  • Hoang Minh NguyenEmail author
  • Gaurav Kalra
  • Tae Joon Jun
  • Sungpil Woo
  • Daeyoung Kim


Workload prediction is an essential prerequisite to allocate resources efficiently and maintain service level agreements in cloud computing environment. However, the best solution for a prediction task may not be a single model due to the challenge of varied characteristics of different systems. Thus, in this work, we propose an ensemble model, namely ESNemble, based on echo state network (ESN) for workload time series forecasting. ESNemble consists of four main steps, including features selection using ESN reservoirs, dimensionality reduction using kernel principal component analysis, features aggregation using matrices concatenation, and regression using least absolute shrinkage and selection operator for final predictions. In addition, necessary hyperparameters for ESNemble are optimized using genetic algorithm. For experimental evaluation, we have used ESNemble to combine five different prediction algorithms on three recent logs extracted from real-world web servers. Through our experimental results, we have shown that ESNemble outperforms all component models in terms of accuracy and resource allocation and presented the running time of our model to show the feasibility of our model in real-world applications.


Ensemble Echo state network Prediction Web applications Cloud computing 



This research was supported by International Research and Development Program of the National Research Foundation of Korea (NRF) funded by the Ministry of Science, ICT and Future Planning of Korea (2016K1A3A7A03952054), and Smart City R&D project of the Korea Agency for Infrastructure Technology Advancement (KAIA) grant funded by the Ministry of Land, Infrastructure and Transport (MOLIT), Ministry of Science and ICT (MSIT) (Grant 18NSPS-B149386-01).


  1. 1.
    Ali-Eldin A, Kihl M, Tordsson J, Elmroth E (2012) Efficient provisioning of bursty scientific workloads on the cloud using adaptive elasticity control. In: Proceedings of the 3rd Workshop on Scientific Cloud Computing. ACM, pp 31–40Google Scholar
  2. 2.
    Barbeau M, Kranakis E (2007) Principles of ad-hoc networking. Wiley, HobokenCrossRefGoogle Scholar
  3. 3.
    Breiman L (1996) Bagging predictors. Mach Learn 24(2):123–140zbMATHGoogle Scholar
  4. 4.
    CRAN package download logs (2017) Accessed Nov 2018
  5. 5.
    Dutreilh X, Moreau A, Malenfant J, Rivierre N, Truck I (2010) From data center resource allocation to control theory and back. In: 2010 IEEE 3rd International Conference on Cloud Computing (CLOUD). IEEE, pp 410–417Google Scholar
  6. 6.
    EDGAR Log File Data Set (2017) Accessed Nov 2018
  7. 7.
    Erlang AK (1917) Solution of some problems in the theory of probabilities of significance in automatic telephone exchanges. Post Off Electr Eng J 10:189–197Google Scholar
  8. 8.
    Friedman JH (2001) Greedy function approximation: a gradient boosting machine. Ann Stat 29:1189–1232MathSciNetCrossRefzbMATHGoogle Scholar
  9. 9.
    Han R, Guo L, Ghanem MM, Guo Y (2012) Lightweight resource scaling for cloud applications. In: 2012 12th IEEE/ACM International Symposium on Cluster, Cloud and Grid Computing (CCGrid). IEEE, pp 644–651Google Scholar
  10. 10.
    Hasan MZ, Magana E, Clemm A, Tucker L, Gudreddi SLD (2012) Integrated and autonomic cloud resource scaling. In: 2012 IEEE Network Operations and Management Symposium (NOMS). IEEE, pp 1327–1334Google Scholar
  11. 11.
    Hyndman RJ, Khandakar Y et al (2007) Automatic time series for forecasting: the forecast package for R. 6/07. Monash University, Department of Econometrics and Business Statistics, ClaytonGoogle Scholar
  12. 12.
    Jaeger H (2001) The echo state approach to analysing and training recurrent neural networks-with an erratum note. Bonn, Germany: German National Research Center for Information Technology GMD Technical Report 148(34):13Google Scholar
  13. 13.
    Jaeger H, Haas H (2004) Harnessing nonlinearity: predicting chaotic systems and saving energy in wireless communication. Science 304(5667):78–80CrossRefGoogle Scholar
  14. 14.
    Kendall DG (1953) Stochastic processes occurring in the theory of queues and their analysis by the method of the imbedded markov chain. Ann Math Stat 24:338–354MathSciNetCrossRefzbMATHGoogle Scholar
  15. 15.
    Kyoto Traffic Data from Kyoto University’s Honeypots (2017) Accessed Nov 2018
  16. 16.
    Lorido-Botrán T, Miguel-Alonso J, Lozano JA (2012) Auto-scaling techniques for elastic applications in cloud environments. Department of Computer Architecture and Technology, University of Basque Country, Tech Rep EHU-KAT-IK-09-12Google Scholar
  17. 17.
    Lorido-Botran T, Miguel-Alonso J, Lozano JA (2014) A review of auto-scaling techniques for elastic applications in cloud environments. J Grid Comput 12(4):559–592CrossRefGoogle Scholar
  18. 18.
    Lukoševičius M, Jaeger H (2009) Reservoir computing approaches to recurrent neural network training. Comput Sci Rev 3(3):127–149CrossRefzbMATHGoogle Scholar
  19. 19.
    Mell P, Grance T et al (2009) The nist definition of cloud computing. Natl inst Stand Technol 53(6):50Google Scholar
  20. 20.
    Messias VR, Estrella JC, Ehlers R, Santana MJ, Santana RC, Reiff-Marganiec S (2016) Combining time series prediction models using genetic algorithm to autoscaling web applications hosted in the cloud infrastructure. Neural Comput Appl 27(8):2383–2406CrossRefGoogle Scholar
  21. 21.
    Miller M (2008) Cloud computing: web-based applications that change the way you work and collaborate online. Que Publishing, LondonGoogle Scholar
  22. 22.
    Mitchell TM (1997) Machine learning, 1st edn. McGraw-Hill Inc., New York, USAzbMATHGoogle Scholar
  23. 23.
    RightScale Cloud Management (2017) Accessed Nov 2018
  24. 24.
    Sakasegawa H (1977) An approximation formula l \(q\simeq \alpha \cdot \rho \beta /(1-\rho )\). Ann Inst Stat Math 29(1):67–75MathSciNetCrossRefzbMATHGoogle Scholar
  25. 25.
    Schapire RE, Freund Y (2012) Boosting: foundations and algorithms. MIT Press, CambridgezbMATHGoogle Scholar
  26. 26.
    Sollich P, Krogh A (1996) Learning with ensembles: how overfitting can be useful. In: Advances in neural information processing systems (NIPS), vol 8, pp 190–196Google Scholar
  27. 27.
    Urgaonkar B, Shenoy P, Chandra A, Goyal P, Wood T (2008) Agile dynamic provisioning of multi-tier internet applications. ACM Trans Auton Adapt Syst (TAAS) 3(1):1CrossRefGoogle Scholar
  28. 28.
    Watkins CJ, Dayan P (1992) Q-learning. Mach Learn 8(3–4):279–292zbMATHGoogle Scholar

Copyright information

© Springer Science+Business Media, LLC, part of Springer Nature 2019

Authors and Affiliations

  1. 1.School of ComputingKorea Advanced Institute of Science and Technology (KAIST)DaejeonSouth Korea

Personalised recommendations