作者投稿和查稿 主编审稿 专家审稿 编委审稿 远程编辑

计算机工程 ›› 2019, Vol. 45 ›› Issue (4): 307-310. doi: 10.19678/j.issn.1000-3428.0052193

• 开发研究与工程应用 • 上一篇    下一篇

基于对抗自编码网络的水利数据补全方法

季琳雅,吕鑫,陶飞飞,曾涛   

  1. 河海大学 计算机与信息学院,南京 211100
  • 收稿日期:2018-07-24 出版日期:2019-04-15 发布日期:2019-04-15
  • 作者简介:季琳雅(1995—),女,硕士研究生,主研方向为深度学习;吕鑫,讲师、博士;陶飞飞,副教授、博士;曾涛,博士研究生。
  • 基金资助:

    国家重点研发计划(2016YFC0400910)。

Water Conservancy Data Completion Method Based on Adversarial Autoencoders Network

JI Linya,L Xin,TAO Feifei,ZENG Tao   

  1. Department of Computer and Information,Hohai University,Nanjing 211100,China
  • Received:2018-07-24 Online:2019-04-15 Published:2019-04-15

摘要:

在大规模监测系统中,监测点失效会导致数据缺失,从而影响数据分析结果的准确性。为此,提出一种对抗自编码的水利数据补全方法。利用自编码器构造生成网络,并提取监测点的数据特征,将其与训练好的判别网络进行对抗,最终补全待修复的监测数据。实验结果表明,与基于图正则化局部子表示方法相比,该方法具有较高的补全精确度,且均方误差较小,能够有效地重构监测数据。

关键词: 水利监测数据, 数据缺失与补全, 对抗自编码网络, 对抗正则化, 重构误差

Abstract:

In large-scale monitoring systems,failure of monitoring points can lead to data loss,which affects the accuracy of data analysis results.Therefoer,a method for complementing water conservancy data against self-encoding is proposed.The self-encoder is used to construct the network to extract the data features of the monitoring points,and then compete with the trained discriminant network to finally complete the monitoring data to be repaired.The experimental results show that compared with the graph-based regularized local sub-representation method,the proposed method has higher accuracy and less mean square error,which can effectively reconstruct the monitoring data.

Key words: water conservancy monitoring data, data missing and completion, Adversarial Autoencoders(AAE) network, adversarial regularization, reconstruction error

中图分类号: