摘要
在基于流形正则化的半监督极限学习机(SS-ELM)的基础上,利用分块矩阵的运算法则,提出了在线半监督极限学习机(OSS-ELM)方法.为避免在实时学习的过程中由于数据累积引起的内存不足,通过对SS-ELM的目标函数的流形正则项的近似,给出了OSS-ELM的近似算法OSSELM(buffer).在Abalone数据集上的实验显示,OSS-ELM(buffer)在线学习的累计时间与所处理的样本个数呈线性关系,同时,9个公共数据集上的实验表明,OSS-ELM(buffer)的泛化能力与SS-ELM的泛化能力的相对偏差在1%以下.这些实验结果说明,OSS-ELM(buffer)不仅解决了内存问题,还在基本保持SS-ELM泛化能力的基础上大幅度提高了在线学习速度,可以有效应用于在线半监督学习当中.
In this paper,with the help of the rules of block matrix multiplication,an online semi-supervised extreme learning machine（OSS-ELM）was proposed according to semi-supervised extreme learning machine（SS-ELM）based on manifold regularization.By the analysis of the manifoldregularization term of the objective function of SS-ELM,a kind of approximation algorithm of OSS-ELM named OSS-ELM（buffer）was proposed to avoid running out of memory in the process of online learning.The linear relationship between the sample number and the cumulative running time of the OSS-ELM（buffer）was revealed in the experiments using Abalone and the relative deviation of the generalization ability of the OSS-ELM and the SS-ELM is less than 1%in 9public data sets,which show that the OSS-ELM（buffer）not only solves the problem of limited memory,but also improves the speed of online learning while keeping the generalization ability of SS-ELM.This proves that the OSS-ELM（buffer）can be effectively applied to online semi-supervised learning.
出处
《上海交通大学学报》
EI
CAS
CSCD
北大核心
2015年第8期1153-1158,1167,共7页
Journal of Shanghai Jiaotong University
基金
2014年度公益性行业(气象)科研专项(GYHY201406004)
天津市面上基金项目(14JCYBJC21800)资助
关键词
极限学习机
半监督学习
在线学习
流形正则化
extreme learning machine（ELM）
semi-supervised learning
online learning
manifold regularization
