基于分层筛选和动态更新的并行选择集成算法  

Selective Ensemble Learning Algorithm Based on Hierarchical Selection and Dynamic Updating in Parallel

作　　者：吴梅红[1] 郭佳盛 鞠颖[1] 林子雨[1] 邹权[1,2] WU Mei-hong GUO Jia-sheng JU Ying LIN Zi-yu ZOU Quan(Department of Computer Science, Xiamen University,Xiamen 361005,China School of Computer Science and Technology, Tianjin University, Tianjin 300072, China)

机构地区：[1]厦门大学计算机科学系,厦门361005 [2]天津大学计算机科学与技术学院,天津300072

出　　处：《计算机科学》2017年第1期48-52,共5页Computer Science

基　　金：国家自然科学基金(61370010;61303004;31200769)资助

摘　　要：提出一种选择性集成学习算法,该算法利用多线程并行优化基分类器的参数,通过多层筛选和动态更新筛选信息获取最优的候选基分类器集合,解决了以往在集成学习中选择分类器效率低下的问题。集成分类器采用分解合并的策略进行加权投票,通过使用二分法将大数据集的投票任务递归分解成多个子任务,并行运行子任务后合并投票结果以缩短集成分类器的投票运行时间。实验结果表明,相对于传统方法,所提出的算法在平均精度、F1-Measure以及AUC指标上都有着显著提升。In this paper, a selective ensemble learning algorithm was proposed based on hierarchical selection and dy- namic updating, which can optimize the parameters of classifier with multi-thread technique and select the sub sequence set of classifiers based on hierarchical selection and dynamical information. It can solve the problem in the past for choo- sing classifier to ensemble learning inefficiently. In addition, divide-and-conquer strategy is employed to reduce the time cost for ensemble voting. The big voting task can be divided recursively into small child task by dichotomy, then the tasks are executed in parallel and it would conquer the voting result. Experimental results show that the selective algo- rithm can outperform the traditional classification algorithms on F1-Measure and AUC.

关 键 词：选择性集成学习 分治算法 并行计算 分类 

分 类 号：TB183[一般工业技术]