Soft Computing

, Volume 22, Issue 2, pp 511–522 | Cite as

A new dynamic weighted majority control chart for data streams

Methodologies and Application
  • 130 Downloads

Abstract

Dynamics are fundamental properties of batch learning processes. Recently, monitoring dynamic processes has interested many researchers due to the importance of dealing with time-changing data stream processes in real-world applications. In this article, a dynamic weighted majority (DWM)-based identification model is proposed for monitoring small, large as well as covariate shifts in nonstationary processes. The proposed method applies DWM ensemble method to aggregate decisions of different control charts to improve single charts’ performances and to reduce the risk of choosing a nonadequate chart. Also in order to improve the shift adaptation mode, a prediction of class label is used to help in classifying the shift during the changing of the process toward the approximated right direction. The new proposed ensemble chart has the ability to deal with complex datasets and presents a concrete shift identification method based on a classification learning technique of changes in nonstationary processes.

Keywords

Control charts Ensemble methods Concept drift Dynamic learning 

Notes

Acknowledgments

This study was funded by the Deutsher Akademischer Austauschdienst (DAAD) with the Grant No. 91526665 and by a research employment at the Technische Universität Dortmund.

Compliance with ethical standards

Conflict of interest

Authors declare that they have no conflict of interest.

Ethical approval

This article does not contain any studies with human participants or animals performed by any of the authors.

References

  1. Asensio AS, Puig AO, Golobardes E (2014) Robust on line neural learning classifier system for data stream classification tasks. J Soft Comput 18(8):1441–1461CrossRefGoogle Scholar
  2. Butz MV (1995) Rule-based evolutionary online learning systems a principal approach to LCS analysis and design, studies in fuzziness and soft computing. Springer, BerlinGoogle Scholar
  3. Deckert M (2011) Batch weighted ensemble for mining data streams with concept drift. In: Proceedings of the 19th international symposium, foundations of intelligent systems, Poland, pp 290–299Google Scholar
  4. Díaz AO, del Campo-Ávila J, Ramos-Jiménez G, Blanco IF, Mota YC, Hechavarría AM, Morales-Bueno R (2014) Fast adapting ensemble: a new algorithm for mining data streams with concept drift. The Scientific World Journal, Article ID 235810 (in press)Google Scholar
  5. Flaig JJ (2014) The Shewhart EWMA automatic control chart. Glob J Res Eng 14(1):21–26Google Scholar
  6. Flaig JJ (2014) Construction of combined charts based on combining functions. Appl Math Sci 8(84):4187–4200Google Scholar
  7. Gama J, Medas P, Castillo G, Rodrigues P (2004) Learning with drift detection. In: SBIA Brazilian symposium on artificial intelligence, Springer Verlag, pp 286–295Google Scholar
  8. Gibbons RD (1999) Use of combined Shewhart CUSUM control charts for ground water monitoring applications. Gr Water 37(5):682–691CrossRefGoogle Scholar
  9. Kolter JZ, Maloof MA (2005) Using additive expert ensembles to cope with concept drift. In: Proceedings of the 22nd international conference on machine learning (ICML 2005), Bonn, Germany, ACM Press, pp 449–456Google Scholar
  10. Kolter JZ, Maloof MA (2007) Dynamic weighted majority: an ensemble method for drifting concepts. J Mach Learn Res 8:2755–2790MATHGoogle Scholar
  11. Liu S, Yamada M, Collier N, Sugiyama M (2013) Change point detection in time series data by relative density ratio estimation. J Neural Netw 43:72–83CrossRefMATHGoogle Scholar
  12. Li PP, Wu X, Hu X (2010) Mining recurring concept drifts with limited labeled streaming data. In: Proceedings of the 2nd Asian conference on machine learning, ACML, Tokyo, Japan, November 8–10, pp 241–252Google Scholar
  13. Lucas James M (1982) Combined Shewhart CUSUM quality control schemes. J Qual Technol 14(2):51–59Google Scholar
  14. Maloof MA, Michalski RS (2004) Incremental learning with partial instance memory. Artif Intell 154:95–126MathSciNetCrossRefMATHGoogle Scholar
  15. Mejri D, Limam M, Weihs C (2015) Monitoring a dynamic weighted majority method based on real datasets with concept drift, Springer International Publishing Switzerland, Wilhelm A (ed.), Analysis of large and complex data, studies in classification, data analysis, and knowledge organization, doi: 10.1007/978-3-319-25226-1-21
  16. Mejri D, Khanchel R, Limam M (2013) An ensemble method for concept drift in nonstationary environment. J Stat Comput Simul 83:1115–1128MathSciNetCrossRefMATHGoogle Scholar
  17. Michalski R (1969) On the quasi minimal solution of the general covering problem. In: Proceedings of the fifth international symposium on information processing, pp 125–128Google Scholar
  18. Nishida K, Yamauchi K, Omori T (2005) ACE: adaptive classifiers ensemble system for concept drifting environments. In: Multiple classifier systems, \(6{th}\) international workshop, MCS 2005, Seaside, CA, USA, June 13–15, Proceedings, pp 176–185Google Scholar
  19. Raza H, Prasad G, Li Y (2013) Dataset shift detection in nonstationary environments using EWMA charts. In: Proceedings of the IEEE international conference on systems, man and cybernetics, pp 3151–3156Google Scholar
  20. Raza H, Prasad G, Li Y (2015) EWMA model based shift detection methods for detecting covariate shifts in non stationary environments. Pattern Recognit 48(3):659–669CrossRefGoogle Scholar
  21. Sampaio ES, Ho LL, de Medeiros PG (2014) A combined \(np_{x}\) \(\overline{X}\) Control Chart to monitor the process mean in Two Stage Sampling. Qual Reliab Eng Int 30(7):1003–1013CrossRefGoogle Scholar
  22. Street WN, Kim Y (2001) A streaming ensemble algorithm (sea) for large-scale classification. In: Proceedings of the seventh ACM SIGKDD international conference on knowledge discovery and data mining, KDD 01, New York, NY, USA, ACM, pp 377–382Google Scholar
  23. Wang H, Fan W, Yu PS, Han J (2003) Mining concept-drifting data streams using ensemble classifiers, In Proceedings of the ninth ACM SIGKDD international conference on knowledge discovery and data mining, KDD 03, New York, NY, USA. ACM, pp 226–235Google Scholar
  24. Wilson SW (1995) Classifier fitness based on accuracy. Evolut Comput 3(2):149–175CrossRefGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2016

Authors and Affiliations

  1. 1.Technische Universität DortmundDortmundGermany
  2. 2.ISGUniversity of TunisTunisTunisia
  3. 3.Dhofar UniversitySalalahOman

Personalised recommendations