The Journal of Korean Institute of Communications and Information Sciences (한국통신학회논문지)

The Korean Institute of Commucations and Information Sciences (한국통신학회)
권호 표지

MEC; A new decision tree generator based on multi-base entropy

다중 엔트로피를 기반으로 하는 새로운 결정 트리 생성기 MEC

  • 전병환 (국립공주대학교 전자계산학과) ;
  • 김재희 (연세대학교 전자공학과)
  • Published : 1997.03.01
Download PDF
⟨ Previous Next ⟩

Abstract

A new decision tree generator MEC is proposed in this paper, which uses the difference of multi-base entropy as a consistent criterion for discretization and selection of attributes. To evaluate the performance of the proposed generator, it is compared to other generators which use criteria based on entropy and adopt different discretization styles. As an experimental result, it is shown that the proposed generator produces the most efficient classifiers, which have the least number of leaves at the same error rate, regardless of whether attribute values constituting the training set are discrete or continuous.

Keywords

References

  1. 연세대학교 본 대학원 전자공학과 박사학위논문 결정 트리 생성에서 속성의 이산화 및 선택을 위한 다중 엔트로피 전병환
  2. Machine Learning v.1 Induction of Decision Trees J. R. Quinlan
  3. Proc. of 5th Int. Conf. on Machine Learning Improved Decision Trees:A Generalized Version of ID3 J. Cheng;U. M. Fayyad;K. B. Irani;Z. Quian
  4. Proc. of 12th National Confernece on Artificial Intelligence Branching on Attribute Values in Decision Tree Generation U. M. Fayyad
  5. Proc. of 10th National Conference on AI The Attribute Selection Problem in Decision Tree Generation U. M. Fayyad;K. B. Irani
  6. Proc. of 13th International Joint Confernece on Artificial Intelligence Decision Trees in Numerical Attribute Spaces T. V. Merckt
  7. European Working Session on Learning On Changing Continuous Attributes into Ordered Discrete Attributes J. Catlett
  8. Applications of Expert Systems Inductive Knowledge Acquisition: A Case Study J. R. Quinlan;P. J. Compton;K. A. Horn;L. Lazarus;Quinlan J. R.
  9. Proc. of 13th International Joint Conference on Artificial Intelligence Multi-Interval Discretization of Continuous-Valued Attributes for Classification Learning U. M. Fayyad;K. B. Irani
  10. Proc. of the 8th National Conference on Artificial Intelligence What Should Be Minimized in a Decision Tree? U. M. Fayyad;K. B. Irani