论文部分内容阅读
提出一种基于病毒协同进化微粒群的最小属性约简算法.在算法中,进化在宿主与病毒种群之间协同进行,通过满足约简分辨力不变条件的最优病毒种子复制操作产生病毒库,病毒通过感染操作在宿主种群完成横向局部搜索,以提高算法局部精确解搜索能力;同时通过删减操作完成自我更新,实现增加局部搜索范围的目的.最后对UCI数据集进行属性约简实验,结果表明该算法在搜索最小属性约简解方面优于其他进化算法,同时收敛速度及寻优效率也有较大提高.
This paper proposes a minimal attribute reduction algorithm based on co-evolutionary particle swarm optimization algorithm, in which evolution proceeds collaboratively between host and virus population, and generates virus database by the optimal virus seed copy operation , The virus completes the horizontal local search in the host population through the infection operation to improve the local exact solution search capability of the algorithm and completes the self-renewal through the deletion operation to achieve the purpose of increasing the local search range.Finally, the attribute reduction experiment is performed on the UCI data set, The results show that the proposed algorithm is superior to other evolutionary algorithms in searching for the minimum attribute reductions, and the convergence speed and optimization efficiency are greatly improved.