Novel ways of Improving Accuracy and Performance in Ensemble Classifiers with Multiple Unbalanced Data

International Journal of Computer Science and Engineering
© 2014 by SSRG - IJCSE Journal
Volume 1 Issue 8
Year of Publication : 2014
Authors : Praveena Prabakaran

pdf
Citation:
MLA Style:

Praveena Prabakaran, "Novel ways of Improving Accuracy and Performance in Ensemble Classifiers with Multiple Unbalanced Data" SSRG International Journal of Computer Science and Engineering 1.8 (2014): 1-5.

APA Style:

Praveena Prabakaran, (2014). Novel ways of Improving Accuracy and Performance in Ensemble Classifiers with Multiple Unbalanced Data. SSRG International Journal of Computer Science and Engineering 1.8, 1-5.

Abstract:

Imbalance classification problem is considered to be one of the emergent challenges in machine learning algorithm. This problem occurs when the number of examples that represents one of the classes of the dataset is much lower than the other classes. A multi objective genetic programming approach to evolving accurate and diverse ensembles of genetic program classifiers with good performance on both the minority and majority of classes. Six benchmark binary classification problems are taken in the existing work. The main objective of the proposed work multiclass datasets are taken to improve the accuracy of minority class and two classes can be classified and each majority and minority class has specified value. The two popular Pareto-based fitness schemes in the multi objective genetic programming algorithm, SPEA2 and NSGAII can be effective in evolving a good set of non dominated solutions in some tasks, this performance needs to be improved for difficult classification problems. The importance of developing an effective fitness evaluation strategy in the underlying MOGP algorithm to evolve good ensemble members.

References:

[1] Genetic Programming for Classification With Unbalanced Data,”IEEE transactions on evolutionary computation, vol. 17, no. 3, 2013.
[2] A.Mclntyre and M.Heywood, “Classification as clustering: A Pareto cooperative-competitive GP approach,” Evol. Comput., vol. 19, no. 1, pp. 137–166, 2011. .
[3] U.Bhowan, M.Zhang, and M.Johnston, “Genetic programming for classification with unbalanced data,” in Proc. 13th Eur. Conf. Genet.Programming, LNCS 6021. 2010. .
[4] U.Bhowan, M.Johnston, and M.Zhang, “Multiobjective genetic programming for classification with unbalanced data,” in Proc. 22nd Australasian Joint Conf. Artif. Intell., LNCS 5866. 2009, pp. 370–380. .
[5] S.Wang, K.Tang, and X.Yao, “Diversity exploration and negative correlation learning on imbalanced data sets,” in Proc. Int. Joint Conf. Neural Netw., 2009, pp. 3259–3266.
[6] N.Chawla and J.Sylvester, “Exploiting diversity in ensembles: Improving the performance on unbalanced datasets,” in Proc. 7th Int. Conf. MCS, 2007, pp. 397–406. .
[7] A.Mclntyre and M.Heywood, “Multiobjective competitive coevolution for efficient GP classifier problem decomposition,” in Proc. IEEE Int. Conf. Syst., Man, Cybern., Oct. 2007, pp. 1930–1937. .
[8] E.Alfaro-Cid, K.Sharman, and A.Esparcia-Alcazar, “A genetic programming approach for bankruptcy prediction using a highly unbalanced database,” in Applications of Evolutionary Computing (LNCS, vol. 4448), M. Giacobini, Ed. Berlin, Germany: Springer, 2007, pp. 169– 178. .
[9] C.Coello Coello, G.Lamont, and D.Veldhuizen, Evolutionary Algorithms for Solving Multi-Objective Problems (Genetic and Evolutionary Computation Series). Berlin, Germany: Springer, 2007. .
[10] A.Chandra and X.Yao, “Ensemble learning using multiobjective evolutionary algorithms,” J. Math. Modelling Algorithms, vol. 5, no. 4, pp. 417–445, 2006. .
[11] G.Batista, R.C.Prati, and M.C.Monard, “Balancing strategies and class overlapping,” in Proc. 6th Int. Adv. IDA, LNCS 3646. 2005, pp. 24– 35. .
[12] N.Japcowicz and S.Stephen, “The class imbalance problem: A systematic study,” Intell. Data Anal., vol. 6, no. 5, pp. 429–450, 2002. .
[13] E.Zitzler, M.Laumanns, and L.Thiele, “Spea2: Improving the strength Pareto evolutionary algorithm for multiobjective optimization,” Dept. Electr. Eng., Swiss Federal Instit. Technol., Zurich, Switzerland, TIK Rep. 103, 2001. .
[14] M.Brameier and W.Banzhaf, “Evolving teams of predictors with linear genetic programming,” Genet. Programming Evolvable Mach., vol. 2, no. 4, pp. 381–407, 2001. .
[15] X.Yao and Y.Liu, “Making use of population information in evolutionary artificial neural networks,” IEEE Trans. Syst., Man, Cybern. B, Cybern., vol. 28, no. 3, pp. 417–425, Jun. 1998. [16] Y.Liu and X.Yao, “Negatively correlated neural networks can produce best ensembles,” Australian J. Intell. Inform. Process. Syst., vol. 4, nos. 3– 4, pp. 176–185, 1997. .
[17] D.W.Opitz and J.W.Shavlik, “Generating accurate and diverse members of a neural-network ensemble,” in Advances in Neural Information Processing Systems. Cambridge, MA: MIT Press, 1996, pp. 535–541.
[18] C.Gathercole and P.Ross, “Dynamic training subset selection for supervised learning in genetic programming,” in Proc. 3rd PPSN, LNCsS 866. 1994, pp. 312–321.

Key Words:

CLASSIFICATION, CLASS IMBALANCE LEARNING,GENETICPROGRAMMING, MULTI-OBJECTIVE MACHINE LEARNING.