A Novel Ensemble Method for Regression via Classification Problems | Science Publications

Journal of Computer Science

Amir Ahmad¹  Sami M. Halawani¹  Ibrahim A. Albidewi¹ 

关键词: Regression via Classification (RvC);    ERD ensembles;    classification problem;    decision trees;    Extreme Randomized Discretization (ERD);    Monothetic Contrast Criteria (MCC);    RvC perform;    Mean Square Error (MSE);    neural network;   

DOI  :  10.3844/jcssp.2011.387.393

学科分类：计算机科学（综合）

来源: Science Publications

【 摘 要 】

Problem statement: Regression via Classification (RvC) is a method in which a regressionproblem is converted into a classification problem. A discretization process is used to covertcontinuous target value to classes. The discretized data can be used with classifiers as a classificationproblem. Approach: In this study, we use a discretization method, Extreme RandomizedDiscretization (ERD), in which bin boundaries are created randomly to create ensembles. Results: Weshow that the proposed ensemble method is useful for RvC problems. We show theoretically that theproposed ensembles for RvC perform better than RvC with the equal-width discretization method. Wealso show the superiority of the proposed ensemble method experimentally. Experimental resultssuggest that the proposed ensembles perform competitively to the method developed specifically forregression problems. Conclusion: As the proposed method is independent of the choice of theclassifier, various classifiers can be used with the proposed method to solve the regression method.

【 授权许可】

Unknown