摘要
提出了利用改进的克隆选择算法发现模糊规则的方法。在该方法中,对规则的评价函数不仅包含规则本身的置信度和蕴涵隶属度等特性,也包含表明规则对规则集整体性能影响程度的量化特性,即一致性贡献和完备性贡献。将该方法用于发现股票20日移动平均线与历史量价之间的模糊规则的仿真试验收到了满意结果。
This paper proposes a new method to find fuzzy rules using an improved clonal selection algorithm. In the new method, the evaluation function of rules includes not only the confidence degree and implication membership function, but also the consistency contribution degree and completeness contribute degree which can quantify the impact of one rule to integrality of the rule set. In the demonstration, it uses the method above to find the fuzzy rules between the 20 days moving average line and other history price variables, and get content results.
出处
《计算机工程》
CAS
CSCD
北大核心
2007年第10期199-201,共3页
Computer Engineering
关键词
克隆选择算法
小生境克隆选择算法
模糊规则
规则发现
股票预测
Clonal selection algorithm
Niche clonal selection algorithm
Fuzzy rule
Rule discovery
Stock forecast
