计算机工程 ›› 2020, Vol. 46 ›› Issue (6): 314-320.doi: 10.19678/j.issn.1000-3428.0055420

• 开发研究与工程应用 • 上一篇    

一种适用于物联网的在线GP-ELM算法

张杰, 沈苏彬   

  1. 南京邮电大学 计算机学院, 南京 210046
  • 收稿日期:2019-07-08 修回日期:2019-09-18 发布日期:2019-09-25
  • 作者简介:张杰(1994-),男,硕士,主研方向为数据挖掘;沈苏彬,研究员、博士。
  • 基金项目:
    江苏省未来网络前瞻性研究项目(BY20130951108)。

An Online GP-ELM Algorithm Suitable for Internet of Things

ZHANG Jie, SHEN Subin   

  1. School of Computer Science, Nanjing University of Posts and Telecommunications, Nanjing 210046, China
  • Received:2019-07-08 Revised:2019-09-18 Published:2019-09-25

摘要: 为满足物联网环境下边缘设备对机器学习算法准确、快速以及自适应产生参数的需求,在DE-ELM的基础上提出一种在线的GP-ELM算法。通过改进结点增加方式,在每次增加结点的同时添加结点统计和结点删除步骤,提高训练速度,同时保持算法的准确性。运用Matlab软件对图片分割、卫星图片分类、卫星DNA等数据集进行训练实验,结果表明,与EI-ELM、D-ELM、EM-ELM等算法相比,GP-ELM算法在准确率、训练时间、模型大小和泛化能力等方面都表现出较好的学习性能。

关键词: 极限学习机, 动态学习, 物联网, 机器学习, 剪枝

Abstract: In order to meet the requirements of edge devices for accurate,fast and adaptive generated parameters of machine learning in the Internet of Things(IoT),this paper proposes an online GP-ELM algorithm based on the differential-evolution extreme learning machine,this paper proposes an online GP-ELM algorithm.The algorithm improves the way of adding nodes by carrying on node statistics and deleting nodes while everytime a node is added,so as to increase the training speed,while maintaining the accuracy of the algorithm.Matlab software is used to train the dataset image segmentation,satellite image classification,satellite DNA and conduct experiments.Results show that compared with EI-ELM,D-ELM,EM-ELM and other algorithms,GP-ELM algorithm has better performance in accuracy,training time,model size and generalization ability.

Key words: Extreme Learning Machine(ELM), dynamic learning, Internet of Things(IoT), machine learning, pruning

中图分类号: