蚁群优化聚类算法在企业效率提升方面的应用研究
2020-09-21徐菁鸿臧英杰
徐菁鸿 臧英杰



摘 要: 常规的聚类算法由于不具备全局性的分析能力,导致企业效率提升效果较差。为此,研究蚁群优化聚类算法在企业效率提升方面的应用。该算法在论域空间建立最大近邻粗糙集,获取影响企业效率提升的因素基本特征;通过设置蚁群算法二维网格定义相似度参数,计算数据的捡拾概率和遗弃概率;根据算法的4个聚类阶段,实现对企业资源的合理分配,从而提升企业效率。实验结果表明:与3种常规方法相比,所提算法具有极强的全局性分析能力,因而在企业效率提升方面的应用效果更佳。由此可见,蚁群优化聚类算法的应用更能满足企业的发展现状。
关键词: 蚁群优化聚类算法; 企业效率; 粗糙集; 相似度; 捡拾概率; 遗弃概率; 聚类与分配
中图分类号: TN911.1?34; TP391.4 文献标识码: A 文章编号: 1004?373X(2020)17?0151?04
Abstract: The conventional clustering algorithms lack the ability of global analysis, which leads to poor effect of efficiency improvement. Therefore, the application of ant colony optimization clustering algorithm in the improvement of enterprise efficiency is studied. In the proposed algorithm, the largest neighbor rough set is built in the universe to obtain the basic characteristics of the factors that affect the improvement of enterprise efficiency. The two?dimensional grid of ant colony algorithm is set to define the similarity parameters for the calculation of picking?up probability and abandonment probability of data. According to the 4 clustering stages of the algorithm, the rational allocation of enterprise resources is realized, thus the enterprise efficiency is improved. The experimental results show that, in comparison with the 3 conventional methods, the proposed algorithm has powerful global analysis ability, so it has a better application effect in the improvement of enterprise efficiency. Thus it can be seen that the ant colony optimization clustering algorithm can better meet the development of enterprise.
Keywords: ant colony optimization clustering algorithm; enterprise efficiency; rough set; similarity; picking?up probability; abandonment probability; clustering and allocation
0 引 言
随着科技手段的不断成熟,越来越多的企业走向系统化、一体化的发展方向,形成更具有自身特色的产业结构链。企业生产运作的最终目的就是获取最大的经济效益,通过丰厚的收入优化企业生产与运营方式,建立一个高效的生產体系。由于企业效益与效率之间是相互依存、相互支撑的关系,因此,为保证企业经济效益,文献[1]提出基于数据挖掘的算法,通过模糊分析影响企业效率因素,提升企业效率。文献[2]结合影响因素熵,通过多种群博弈蚁群算法,找出企业效益与效率之间的关系,以此实现效率提升。文献[3]则利用蚁群算法的多条路径,类比分析历史影响因素特征,得出可以提升企业效率的参数。……
