| 一种基于模糊聚类的可解释性建模方法 |
| |
| 引用本文： | 王晓兰,曾贤强,王文琰.一种基于模糊聚类的可解释性建模方法[J].甘肃科学学报,2006,18(4):75-79. |
| |
| 作者姓名： | 王晓兰 曾贤强 王文琰 |
| |
| 作者单位： | 1. 兰州理工大学,电气工程与信息工程学院,甘肃,兰州,730000
2. 兰州理工大学,机械工厂,甘肃,兰州,730050 |
| |
| 摘 要： | 提出一种基于模糊聚类的可解释性建模方法.利用提出的一种含有熵的聚类有效性函数来评价模糊聚类方法的有效性和可解释性,从而确定模糊规则数和模型前提参数,然后利用最小二乘法来辨识模型的结论参数,最后采用梯度下降法来调整模型的参数.该方法应用于Box-Jenkins 数据仿真实例,仿真结果表明该方法不但能保证系统的精确性,还具有很高的可解释性.
|
| 关 键 词： | 模糊建模 解释性 模糊聚类 熵 |
| 文章编号： | 1004-0366（2006）04-0075-05 |
| 收稿时间： | 2005-11-11 |
| 修稿时间： | 2005年11月11 |
A Method of Interpretable Modeling Based on Fuzzy Clustering |
| |
| WANG Xiao-lan,ZENG Xian-qiang,WANG Wen-yan.A Method of Interpretable Modeling Based on Fuzzy Clustering[J].Journal of Gansu Sciences,2006,18(4):75-79. |
| |
| Authors: | WANG Xiao-lan ZENG Xian-qiang WANG Wen-yan |
| |
| Institution: | 1. School of Electrical Engineering and Information Engineering, Lanzhou University of Science and Technology, Lanzhou 730000, China;2. Machine Work-shop, Lanzhou University of Science and Technology, Lanzhou 730050, China |
| |
| Abstract: | A method of interpretable modeling based on fuzzy clustering is proposed. An entropy-based fuzzy partition validity index is used to improve the interpretability of fuzzy models. The fuzzy clustering technique associated with the proposed validity index is used to find out the optimal number of fuzzy rules and the membership functions in the antecedent part, and the parameters in the consequent part are found out by means of the least square algorithm adjust parameters of fuzzy model. Finally, Box-Jenkins data set. It is verified that interpretable. , and then the gradient descent algorithm is used to precisely the method is successfully applied to identify the well-known the method is computationally efficient and linguistically |
| |
| Keywords: | fuzzy modeling interpretability fuzzy clustering entropy |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
