融合基于MapReduce并行改进二元蚁群算法与分形维数的属性选择方法  

Attribute Selection Method Combined with MapReduce-Based Improved BACO and Fractal Dimension

作　　者：许力分 倪志伟[1,2] 朱旭辉[1,2] 贾凯 伍章俊[1,2] XU Lifen;NI Zhiwei;ZHU Xuhui;JIA Kai;WU Zhangjun(School of Management, Hefei University of Technology, Hefei 230009;Key Laboratory of Process Optimization and Intelligent Decision-Making, Hefei University of Technology, Hefei 230009)

机构地区：[1]合肥工业大学管理学院,合肥230009 [2]合肥工业大学过程优化与智能决策教育部重点实验室,合肥230009

出　　处：《系统科学与数学》2019年第6期918-933,共16页Journal of Systems Science and Mathematical Sciences

基　　金：国家自然科学基金重大研究计划培育项目(91546108),国家自然科学基金重大项目(91490725),国家自然科学基金创新群体项目(71521001);安徽省自然科学基金(1908085QG298);国家重点研发计划(2016YFF0202604);中央高校基本科研业务费专项资金(JZ2019HGTA0053,JZ2019HGBZ0128)资助课题

摘　　要：属性选择是数据挖掘领域用于降低数据特征维度的预处理方法.针对大数据环境下高维数据的属性约简问题,提出了融合基于MapReduce并行改进二元蚁群算法与分形维数的属性选择方法.首先,引入了参数控制的位置更新策略、对蚂蚁个体与种群进行交叉变异、重新定义阻塞机制的信息素更新,提出了并行改进的二元蚁群算法MRIBACO.其次,以并行二元蚁群算法作为离散解空间的搜索策略,结合分形维数提出了属性选择模型.在6个UCI数据集上的实验结果表明,较其他方法计算效率更优,同时表明了其有效性与稳定性.Attribute selection is a primary preprocessing step for reducing dimension of datasets in data mining.When it comes to an attribute selection problem of high-dimensional data under big data environment,an attribute selection method based on an improved Binary Ant Colony Optimization Algorithm(MRIBACO)and fractal dimension is proposed.Firstly,an improved Binary Ant Colony Optimization Algorithm is proposed,combining with MapReduce programming model,by introducing an ariable parameter selection of location strategy,a cross variation strategy for partial optimization and a new pheromone updating rule with blocking mechanism.Secondly,fractal dimension is used as an evaluation standard for attribute subsets,and MRIBACO algorithm is employed as search strategy in the discrete solution space.The experimental results on 6 UCI datasets show that,compared with other algorithms,this algorithm achieves a better operation efficiency,and its effectiveness and stability are better than others.

关 键 词：属性选择 分形理论 二元蚁群优化算法 MAPREDUCE 分群分治 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]