Abstract
模糊决策树的ID3算法是Quinlan提出的传统ID3算法的一个模糊版本？树的整个产生过程在给定的显著性水平α的基础上进行，的值在很大程度上影响模糊熵的计算，从而影响模糊决策树最终的分类结果？对参数α关于模糊熵的敏感性进行了分析，试图定性地找出二者之间的解析关系，从而为选取参数α的值以达到最优的分类结果提供理论依据？
Fuzzy-ID3 algorithm in the fuzzy decision tree is the fuzzy version of the ID3 algorithm proposed by Quinlan in 1986. The process of building trees is based on significant levelα , the value of αaffect the computation of fuzz y entropy and classification result of fuzzy decision tree in large degree. This paper analyzes the sensitivity of parameter αabout fuzzy entropy, expects to find qualitatively the analytic relationships between them, and provides the academic evidence of selecting parameter α in order to gain the best classification result.
Fuzzy-ID3 algorithm in the fuzzy decision tree is the fuzzy version of the ID3 algorithm proposed by Quinlan in 1986. The process of building trees is based on significant levelα , the value of αaffect the computation of fuzz y entropy and classification result of fuzzy decision tree in large degree. This paper analyzes the sensitivity of parameter αabout fuzzy entropy, expects to find qualitatively the analytic relationships between them, and provides the academic evidence of selecting parameter α in order to gain the best classification result.
| Translated title of the contribution | Sensitivity analysis of parameters about fuzzy entropy in fuzzy decision tree |
|---|---|
| Original language | Chinese (Simplified) |
| Pages (from-to) | 90-92 |
| Number of pages | 3 |
| Journal | 计算机工程 = Computer Engineering |
| Volume | 29 |
| Issue number | 11 |
| Publication status | Published - 2003 |
| Externally published | Yes |
Bibliographical note
基金: 河北省自然科学基金资助项目 (698139)Keywords
- 归纳学习
- 模糊熵
- 模糊决策树
- Fuzzy decision tree
- Fuzzy entropy
- Inductive learning