论文部分内容阅读
属性约简与规则分类学习是粗糙集理论研究和应用的重要内容.文中充分利用量子计算加速算法速度和混合蛙跳算法高效协同搜索等优势,提出一种基于动态交叉协同的量子蛙跳属性约简与分类学习的级联算法.该算法用量子态比特进行蛙群个体编码,以动态量子角旋转调整策略实现属性染色体快速约简,并在粗糙熵阈值分类标准内采用量子蛙群混合交叉协同进化机制提取和约简分类规则、组合决策规则链等,最后构造属性约简和分类学习双重功能级联模型.仿真实验验证该算法不仅具有较高的全局优化性能,且属性约简与规则分类学习的精度和效率均超过同类算法.
Attribute reduction and rule classification are the important contents of research and application of rough set theory.This paper makes full use of the advantages of quantum computing acceleration algorithm and hybrid frog leaping algorithm efficient collaborative search, and proposes a quantum leap frog attribute based on dynamic cross-cooperation Concise and classification learning.The algorithm encodes individuals of frog with quantum bits, and realizes the rapid reduction of attribute chromosomes by using dynamic quantum rotation control strategy, and adopts quantum frog swarm hybrid cross-cooperation Evolutionary mechanism to extract and reduce classification rules, combined decision rules chain, and finally construct a dual-function cascading model of attribute reduction and classification learning.Experimental results show that the proposed algorithm not only has higher global optimization performance, but also reduces attribute and rules classification learning The accuracy and efficiency are more than similar algorithms.