论文部分内容阅读
讨论了最大频繁序列模式和公平竞争层次模型(HFC),设计了最大频繁序列模式的挖掘算法(MFSPMA),把MFSPMA同HFC结合起来,提出了基于序列挖掘技术的分等级搜索可持续进化算法(SEAHSM).该进化算法设置多个不同层次的种群为不同适应度水平的个体提供生存空间,采用最大频繁子模式挖掘算法挖掘种群中的优良基因,并将具有优良基因模块的新个体注入到不同适应度水平的种群,从而实现遗传信息的稳定继承,有效避免优良基因的丢失.实验结果表明:SEAHSM在维持遗传信息稳定性、避免早熟收敛、提高搜索精度等方面表现良好.
The maximum frequent sequence model and the fair competition hierarchy model (HFC) are discussed. The algorithm of the largest frequent sequence pattern mining (MFSPMA) is designed. Combining the MFSPMA with the HFC, a hierarchical mining sustainable evolution algorithm based on sequence mining is proposed SEAHSM). This evolutionary algorithm sets up the living space for individuals with different fitness levels by multiple levels of population, uses the maximal frequent sub-pattern mining algorithm to mine good genes in the population, and injects new individuals with excellent gene modules into different Fitness level, so as to achieve a stable inheritance of genetic information and avoid the loss of good genes effectively.The experimental results show that SEAHSM performs well in maintaining the stability of genetic information, avoiding premature convergence and improving search precision.