摘要
通过定义单边三角形模糊数空间上的一种全序关系,提出了属性取值为单边三角形模糊数的决策树学习算法.作为ID3算法在单边三角形模糊数意义下的推广,算法通过一种分割信息熵的极小化来选取扩展属性.通过非平稳割点的分析,减少了分割信息的计算次数,使算法的效率得到了提高.
The paper defines a total order on one-side triangular fuzzy numbers,then presents rules of learning algorithm,that is,one-side decision trees or rules using the information entropy minimization heuristic.The result serves to give a better understanding of the entropy measure,to point out that the behavior of the information entropy heuristic is desirable,and to improve the efficiency of evaluating attributes,for cutting value selection.
出处
《南华大学学报（自然科学版）》
2007年第3期81-83,88,共4页
Journal of University of South China：Science and Technology
关键词
机器学习
归纳学习
单边三角形模糊数
决策树
类信息熵极小
Machine learning
Induction learning
One-side-Triangle-Fuzzy-Number
Decision tree
information entropy minimization
