Partition Real Data in Decision Tree Using Statistical Criterion

Abstract:

Partition methods for real data play an extremely important role in decision tree algorithms in data mining and machine learning because the decision tree algorithms require that the values of attributes are discrete. In this paper, we propose a novel partition method for real data in decision tree using statistical criterion. This method constructs a statistical criterion to find accurate merging intervals. In addition, we present a heuristic partition algorithm to achieve a desired partition result with the aim to improve the performance of decision tree algorithms. Empirical experiments on UCI real data show that the new algorithm generates a better partition scheme that improves the classification accuracy of C4.5 decision tree than existing algorithms.

• Periodical: Applied Mechanics and Materials (Volumes 380-384)
• Pages: 1469-1472
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.380-384.1469
• Online since: August 2013
• Authors: Gui Jun Shan
• Keywords: C4.5, Decision Tree, Partition, Statistical Criterion

References

[1] H. Liu, F. Hussain, C. L. Tan, and M. Dash, Discretization: an enabling technique, Journal of Data Mining and Knowledge Discovery, vol. 6, no. 4, p.393–423, (2002).

[2] J. Dougherty, R. Kohavi, and M. Sahami, Supervised and unsupervised discretization of continuous feature, Proceedings of 12th International Conference of Machine learning, p.194–202, (1995).

DOI: 10.1016/b978-1-55860-377-6.50032-3

[3] L. A. Kurgan and K. J. Cios, CAIM discretization algorithm, IEEE Transactions on Knowledge and Data Engineering, vol. 16, no. 2, p.145–153, (2004).

DOI: 10.1109/tkde.2004.1269594

[4] J. Catlett, On changing continuous attributes into ordered discrete attributes, In Proceedings of Fifth European Working Session on Learning. Berlin: Springer-Verlag, p.164–177, (1991).

DOI: 10.1007/bfb0017012

[5] E. H. Tay and L. Shen, A modified chi2 algorithm for discretization, IEEE Transactions on Knowledge and Data Engineering, vol. 14, no. 3, p.666–670, (2002).

DOI: 10.1109/tkde.2002.1000349

[6] C. T. Su and J. H. Hsu, An extended chi2 algorithm for discretization of real value attributes, IEEE Transactions on Knowledge and Data Engineering, vol. 17, no. 3, p.437–441, (2005).

DOI: 10.1109/tkde.2005.39

[7] M. Biba, F. Esposito, S. Ferilli, N. D. Mauro, and T. Basile, Unsupervised discretization using kernel density estimation, the Twentieth International Joint Conference on Artificial Intelligence (IJCAI), p.696–701, (2007).

[8] U. Fayyad and K. Irani, Multi-interval discretization of continuous-valued attributes for classification learning, In Proc. Thirteenth International Joint Conference on Artificial Intelligence. San Mateo, CA: Morgan Kaufmann, p.1022–1027, (1993).

[9] A. Bondu, M. Boulle, V. Lemaire, S. Loiseau, and B. Duval, A non-parametric semi-supervised discretization method, 2008 Eighth IEEE International Conference on Data Mining (ICDM), p.53–62, (2008).

DOI: 10.1109/icdm.2008.35

[10] E. Armengol and A. Garcia-Cerdana, Refining discretizations of continuous-valued attributes, Modeling of Decisions of Artificial Intelligence Conference, LNAI, Springer, Heidelberg, p.258–269, (2012).

DOI: 10.1007/978-3-642-34620-0_24

[11] G. Salvador, L. Julian, S. J. Antonio, L. Victoria, and H. Francisco, A survey of discretization techniques: Taxonomy and empirical analysis in supervised learning, IEEE Transactions on Knowledge and Data Engineering, vol. 25, no. 4, p.734–750, (2013).

DOI: 10.1109/tkde.2012.35

[12] S. Hettich and S. D. Bay, The UCI KDD archive [db/ol], http: /kdd. ics. uci. edu/, (1999).