基于属性约简的自采样集成分类方法
作者:
作者单位:

1.重庆邮电大学计算智能重庆市重点实验室,重庆 400065;2.重庆邮电大学计算机科学与技术学院,重庆 400065

作者简介:

通讯作者:

基金项目:

国家自然科学基金(61876027, 61751312, 61533020)资助项目。


Self-sampling Ensemble Classification Method Based on Attribute Reduction
Author:
Affiliation:

1.Chongqing Key Laboratory of Computational Intelligence, Chongqing University of Posts and Telecommunications, Chongqing 400065, China;2.School of Computer Science and Technology, Chongqing University of Posts and Telecommunications, Chongqing 400065, China

Fund Project:

  • 摘要
  • |
  • 图/表
  • |
  • 访问统计
  • |
  • 参考文献
  • |
  • 相似文献
  • |
  • 引证文献
  • |
  • 资源附件
    摘要:

    现有的集成技术大多使用经过训练的各个分类器来组成集成系统,集成系统的庞大导致产生额外的内存开销和计算时间。为了提高集成分类模型的泛化能力和效率,在粗糙集属性约简的研究基础上,提出了一种基于属性约简的自采样集成分类方法。该方法将蚁群优化和属性约简相结合的策略应用在原始特征集上,进而得到多个最优的特征约简子空间,以任意一个约简的特征子集作为集成分类的特征输入,能在一定程度上减少分类器的内存消耗和计算时间;然后结合以样本的学习结果和学习速度为约束条件的自采样方法,迭代训练每个基分类器。最后实验结果验证了本文方法的有效性。

    Abstract:

    The ensemble learning technology often uses each basic classifier that has been trained to form a complete ensemble system, and the largeness of the ensemble system easily leads to more memory and time. So as to gain the high prediction correctness and low classification time of the ensemble classification model, according to the research of attribute reduction in rough sets, this paper proposes a self-sampling ensemble classification method based on the attribute reduction. This method applies the strategy of combining ant colony optimization and attribute reduction to the original feature data set, and then multiple optimal feature reduction subspaces are obtained. Taking any feature subset after reduction as the feature input of the integrated classifier can reduce the memory usage and classification time of the classifier to some extent. And then each self-sampling method taking the learning results and learning speed of the samples as constraints is combined to iteratively train each base classifier. Finally, the feasibility of the proposed method is further proved by experimental results.

    表 3 混淆矩阵Table 3 Confusion matrix
    表 4 UCI数据集信息Table 4 Information of UCI data sets
    图1 基于属性约简的自采样集成分类模型Fig.1 Self-sampling ensemble classification model based on attribute reduction
    图2 SS-AdaBoost与RSS-AdaBoost方法错误率变化Fig.2 Error rate variety of SS-AdaBoost and RSS-AdaBoost algorithms
    图3 原始属性与各约简属性分类效果比较Fig.3 Comparison of classification effects between original attributes and reduced attributes
    表 1 信息表Table 1 Table of information
    表 2 Table 2 Discernibility matrix of
    表 5 属性约简子集Table 5 Attribute reduction subsets
    表 6 实验结果Table 6 Experimental results
    参考文献
    相似文献
    引证文献
引用本文

李朋飞,于洪.基于属性约简的自采样集成分类方法[J].数据采集与处理,2021,36(3):498-508

复制
分享
文章指标
  • 点击次数:
  • 下载次数:
历史
  • 收稿日期:2020-05-23
  • 最后修改日期:2020-11-01
  • 录用日期:
  • 在线发布日期: 2021-06-16