DCR: Discretization using Class Information to Reduce Number of Intervals

Discretization techniques for data set features have received increasing research attention. Results using discretized features are usually more compact, shorter, and accurate than using continuous values. In this paper, an algorithm called Discretization using Class information to Reduce number of intervals (DCR) is proposed. DCR uses both class information and order between attributes to determine the discretization scheme with minimum number of intervals. Attribute discretization order is determined based on information gain of each attribute with respect to the class attribute. The number of intervals is reduced by deleting training data at each step of attribute discretization. Experiments are performed to compare the predictive accuracy and execution time of this algorithm with several well-known algorithms. Results show that discretized features generated by the DCR algorithm contain a smaller number of intervals than other supervised algorithms using less execution time, and the predictive accuracy is as high or higher.

[1]  Usama M. Fayyad,et al.  On the Handling of Continuous-Valued Attributes in Decision Tree Generation , 1992, Machine Learning.

[2]  Djamel A. Zighed,et al.  Discretization of Continuous Features by Resampling , 2008, EGC.

[3]  Alberto Maria Segre,et al.  Programs for Machine Learning , 1994 .

[4]  Zbigniew W. Ras,et al.  Methodologies for Intelligent Systems , 1991, Lecture Notes in Computer Science.

[5]  Caroline Chan,et al.  Determination of quantization intervals in rule based model for dynamic systems , 1991, Conference Proceedings 1991 IEEE International Conference on Systems, Man, and Cybernetics.

[6]  Randy Kerber,et al.  ChiMerge: Discretization of Numeric Attributes , 1992, AAAI.

[7]  Marc Boullé,et al.  MODL: A Bayes optimal discretization method for continuous attributes , 2006, Machine Learning.

[8]  J. Ross Quinlan,et al.  Induction of Decision Trees , 1986, Machine Learning.

[9]  Huan Liu,et al.  Chi2: feature selection and discretization of numeric attributes , 1995, Proceedings of 7th IEEE International Conference on Tools with Artificial Intelligence.

[10]  Peter Clark,et al.  The CN2 Induction Algorithm , 1989, Machine Learning.

[11]  Krzysztof J. Cios,et al.  Hybrid inductive machine learning: an overview of CLIP algorithms , 2002 .

[12]  Lech Polkowski,et al.  Rough Sets in Knowledge Discovery 2 , 1998 .

[13]  Steven Salzberg,et al.  A Weighted Nearest Neighbor Algorithm for Learning with Symbolic Features , 2004, Machine Learning.

[14]  Jason Catlett,et al.  On Changing Continuous Attributes into Ordered Discrete Attributes , 1991, EWSL.

[15]  Kenneth A. Kaufman,et al.  Learning from Inconsistent and Noisy Data: The AQ18 Approach , 1999, ISMIS.

[16]  Andrew K. C. Wong,et al.  Class-Dependent Discretization for Inductive Learning from Continuous and Mixed-Mode Data , 1995, IEEE Trans. Pattern Anal. Mach. Intell..

[17]  Thanh-Nghi Do,et al.  A Comparison of Different Off-Centered Entropies to Deal with Class Imbalance for Decision Trees , 2008, PAKDD.

[18]  J. Ross Quinlan,et al.  C4.5: Programs for Machine Learning , 1992 .

[19]  Paul D. Scott,et al.  Zeta: A Global Method for Discretization of Continuous Variables , 1997, KDD.

[20]  Lukasz A. Kurgan,et al.  CAIM discretization algorithm , 2004, IEEE Transactions on Knowledge and Data Engineering.