| 基于多目标遗传算法的模糊分类系统设计 |
| |
| 引用本文： | 邢宗义,侯远龙,贾利民.基于多目标遗传算法的模糊分类系统设计[J].东南大学学报(自然科学版),2006,36(5):725-731. |
| |
| 作者姓名： | 邢宗义 侯远龙 贾利民 |
| |
| 作者单位： | 南京理工大学机械工程学院,南京,210094;北京交通大学交通运输学院,北京,100044 |
| |
| 基金项目： | 国家自然科学基金资助项目(60332020) |
| |
| 摘 要： | 提出了一种基于多目标遗传算法的模糊建模方法,实现了具备最大分类精度、最少特征变量和模糊规则数目的模糊分类系统的设计.首先,为缓解维数灾难问题,采用多目标遗传算法进行特征变量的选择和论域的模糊划分,构造基于栅格划分的初始模糊分类系统.然后为减少规则数目,提高模糊系统的解释性,采用遗传算法进行规则选择,得到具备较少规则数目的精简模糊分类系统.最后为提高精确性,采用约束遗传算法对精简模糊分类系统进行整体优化,在解释性不变的前提下,提高模糊分类系统的精确性.Iris和W ine分类系统的仿真,验证了该方法的有效性.
|
| 关 键 词： | 模糊分类系统 多目标遗传算法 特征变量选择 模糊划分 |
| 文章编号： | 1001-0505（2006）05-0725-07 |
| 收稿时间： | 03 5 2006 12:00AM |
| 修稿时间： | 2006-03-05 |
Design of multi-objective genetic-based fuzzy classification system |
| |
| Xing Zongyi,Hou Yuanlong,Jia Limin.Design of multi-objective genetic-based fuzzy classification system[J].Journal of Southeast University(Natural Science Edition),2006,36(5):725-731. |
| |
| Authors: | Xing Zongyi Hou Yuanlong Jia Limin |
| |
| Institution: | 1. School of Mechanical Engineering, Nanjing University of Science and Technology, Nanjing 210094;2 School of Traffic and Transportation, Beijing Jiaotong University, Beijing 100044, China |
| |
| Abstract: | An approach based on multi-objective genetic algorithm is proposed to construct fuzzy classification system with maximum precision and minimum number of features and minimum number of fuzzy rules.First,in order to relieve the problem of dimension disaster,a genetic algorithm is used to accomplish feature selection and fuzzy partition with three objectives,thus an initial fuzzy system is obtained.Second by,a genetic algorithm is employed to select fuzzy rules with two objectives to achieve a compact fuzzy system.Third by,a constrained genetic algorithm is utilized to optimize the compact fuzzy system and improve its interpretability,while its precision is preserved.The proposed approach was applied to the Iris and Wine benchmark problems,and the results verify its validity. |
| |
| Keywords: | fuzzy classification system multi-objective genetic algorithm feature selection fuzzy partition |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
