基于模糊扩张矩阵的启发式算法

A Heuristic Algorithm Based on Fuzzy Extension Matrix

扩张矩阵是一种规则归纳的方法。在模糊环境下,模糊扩张矩阵算法根据路径的最小模糊信息熵标准,从示例中归纳产生一组模糊规则。本文引入评价矩阵的概念,并使用近似集合覆盖算法的思想,对上述模糊规则归纳方法做了改进。实验结果表明,与原算法相比,改进后的版本增添了许多优点,比如,得到的规则较少,运算速度较快等。

Extension matrix is an approach to induce rules.Under fuzzy environment ,the fuzzy extension matrix approach can generate a set of fuzzy rules from examples according to the minimum fuzzy entropy criterion of the path.In this paper,the above approach is improved by introducing evaluation matrix and using the idea of approximate set-covering algorithm.The experimental results show that the improved version has more advantages in several aspects than the old one,such as generating fewer rules and running faster.