面向贯序不均衡分类的粒度极限学习机-控制与决策.PDFVIP

第31 卷 第12 期 控 制 与 决 策 2016 年 12 月 Vol. 31 No. 12 Control and Decision Dec. 2016 文章编号: 1001-0920 (2016) 12-2147-08 DOI: 10.13195/j.kzyjc.2015.1257 面向贯序不均衡分类的粒度极限学习机 毛文涛 , 田杨阳, 王金婉 , 何 玲 (河南师范大学a. 计算机与信息工程学院，b. 河南省高校计算 智能与数据挖掘工程技术研究中心，河南新乡453007) 摘 要: 针对现有算法对贯序到达的密度型不均衡数据分类效果不佳的缺陷, 提出一种基于粒度划分的在线贯序极 限学习机算法. 离线阶段，根据数据分布特性对多类样本进行粒度划分, 用粒心代替原有样本, 建立初始模型; 在线 阶段, 根据更新后的分布特性对多类边界数据进行二次粒度划分, 替换原有边界数据, 并动态更新网络权值. 理论分 析证明该算法存在信息损失上界. 实验结果表明, 该算法能有效提高贯序不均衡数据上的整体泛化性能和分类效率. 关键词: 极限学习机；粒度划分；贯序不均衡数据；欠取样 中图分类号: TP181 文献标志码: A Granular extreme learning machine for sequential imbalanced data MAO Wen-tao , TIAN Yang-yang , WANG Jin-wan , HE Ling (a. College of Computer and Information Engineering，b. Computational Intelligence and Data Mining Engineering Technology Research Center of Colleges and Universities in He’nan Province ，He’nan Normal University ，Xinxiang 453007 ，China ．Correspondent ：MAO Wen-tao ，E-mail ：maowt@) Abstract: Aiming at the shortcomings of the present classification algorithms on density-based imbalanced data which are selected sequentially, an online sequential extreme learning machine based on granular division is proposed. In the offline stage, majority class samples are divided by using granularity according to the data distribution property, and the centre of granule is introduced for replacing the samples in this granule. In the online stage, the boundary majority samples are divided again by using granularity according to the ne