Abstract
Concept drift can be considered as a distribution mismatch problem where class distribution changes as a time passes. This problem is commonly found in classification task of data mining. Among the proposed solutions, the cost-based Class Distribution Estimation (CDE) shows the best performance in coping with difference in class distribution between train and test datasets. However there is still some problem, as CDE lost its performance when there is too much change in class distribution. In this paper, CDE-weight is proposed to reduce the impact of high change in class distribution. The idea is to use many models suitable with many class distributions along with dynamic weighting method that adjusts weight of each model according to its class distribution. Experimented results indicate that CDE-Weight methods are able to reduce the impact of misestimating and improve the classifier performance when train and test data are different.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
Alaiz-Rodriguez, R., Guerrero-Curieses, A., Cid-Sueiro, J.: Minimax.: Regret Classifier for Imprecise Class Distributions. Journal of Machine Learning Research 8 (2007)
Alaiz-RodrÃguez, R., Japkowicz, N.: Assessing the Impact of Changing Environments on Classifier Performance. In: Bergler, S. (ed.) Canadian AI. LNCS (LNAI), vol. 5032, pp. 13–24. Springer, Heidelberg (2008)
Asuncion, A., Newman, D.J.: UCI Machine Learning Repository (2007), http://www.ics.uci.edu/~mlearn/MLReposito-ry.html
Chapelle, O., Scholkopf, B., Zien, A.: Semi-Supervised Learning. MIT Press, Cambridge (2006)
Chawla, N.V., Bowyer, K.W., Kegelmeyer, W.P.: SMOTE: Synthetic Minority Over-sampling Technique. Journal of Artificial Intelligence Research 16 (2002)
Elkan, C.: The foundations of cost-sensitive learning. In: The Proceedings of the 17th International Joint Conference on Artificial Intelligence, pp. 973–978 (2001)
Forman, G.: Quantifying counts and costs via classification. Data Mining Knowledge Discovery 17, 2 (2008)
Latinne, P., Saerens, M., Decaestecker, C.: Adjusting the Outputs of a Classifier to New a Priori Probabilities May Significantly Improve Classification Accuracy: Evidence from a multi-class problem in remote sensing. In: The Proceedings of the 18th International Conference on Machine Learning, pp. 298–305 (2001)
Limsetto, N., Waiyamai, K.: Handling Concept Drift via Ensemble and Class Distribution Estimation Technique. In: Tang, J., King, I., Chen, L., Wang, J. (eds.) ADMA 2011, Part II. LNCS, vol. 7121, pp. 13–26. Springer, Heidelberg (2011)
Provost, F., Fawcett, T.: Robust Classification for Imprecise Environments. Machine Learning 42, 3 (2001)
Provost, F., Fawcett, T., Kohavi, R.: The Case against Accuracy Estimation for Comparing Induction Algorithms. In: The Proceedings of the 15th International Conference on Machine Learning, pp. 445–453 (1998)
Quinlan, R.J.: C4.5: Programs for Machine Learning. Morgan Kaufmann (1993)
Rokach, L.: Ensemble-based classifiers. Artificial Intelligence Review 33 (2010)
Saerens, M., Latinne, P., Decaestecker, C.: Adjusting the outputs of a classifier to new a priori probabilities: A simple procedure. Neural Computing 14 (2002)
Tsymbal, A.: The problem of concept drift: Definitions and related work. Computer Science Department, Trinity College Dublin (2004)
Weiss, G.M.: Mining with rarity: a unifying framework. SIGKDD Explorations Newsletter 6, 1 (2004)
Weiss, G.M., Provost, F.: Learning when training data are costly: The effect of class distribution on tree induction. Journal of Artificial Intelligence Research 19 (2003)
Witten, I.H., Frank, E.: Data Mining: Practical Machine Learning Tools and Techniques, 2nd edn. Morgan Kaufmann (2005)
Xue, J.C., Weiss, G.M.: Quantification and semi-supervised classification methods for handling changes in class distribution. In: Proceedings of the 15th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, Paris, France, pp. 897–906. ACM (2009)
Zadrozny, B., Elkan, C.: Learning and making decisions when costs and probabilities are both unknown. In: The Proceedings of the 7th International Conference on Knowledge Discovery and Data Mining, pp. 204–213 (2001)
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2014 Springer International Publishing Switzerland
About this paper
Cite this paper
Limsetto, N., Waiyamai, K. (2014). Integrating Weight with Ensemble to Handle Changes in Class Distribution. In: Perner, P. (eds) Machine Learning and Data Mining in Pattern Recognition. MLDM 2014. Lecture Notes in Computer Science(), vol 8556. Springer, Cham. https://doi.org/10.1007/978-3-319-08979-9_8
Download citation
DOI: https://doi.org/10.1007/978-3-319-08979-9_8
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-08978-2
Online ISBN: 978-3-319-08979-9
eBook Packages: Computer ScienceComputer Science (R0)