期刊文献+

结合特征选择和优化随机森林的无线网络数据丢失重建 被引量:1

Data Reconstruction in Wireless Network Based on Feature Selection and Optimized Random Forests
下载PDF
导出
摘要 基于K邻近(KNN)算法和随机森林算法,提出了一种无线网络中丢失数据的重建方法。首先将多维原始数据通过不稳定无线信道进行发送,接收端将接收到的完整原始数据集中,利用KNN算法筛选出部分和重建特征相关性较高的特征,用于构造随机森林模型。然后输入缺失的数据样本,随机森林模型自适应地对数据样本进行分类,并利用完整样本对缺失特征值进行预测,从而完成丢失数据的重建。最后通过仿真实验表明,该方案可以有效地提升数据重建的精确度,在数据丢失率达到80%的情况下,重建数据的准确率仍然优于现有的解决方案。 A new reconstruction method for lost data is proposed in wireless network by combining K-nearest neighbor(KNN)algorithm feature selection and random forest.Firstly,the multi-dimensional raw data is transmitted through the unstable wireless channel.For the received complete raw feature set,receiver can build random forest model by the features selected by KNN algorithm.These features are considered to be highly correlated with reconstruction features.When the missing data samples are input,the random forest model adaptively classifies the data samples and uses the complete samples to predict the missing feature values.The missing data can finally be reconstructed by predicted feature values.A large number of simulation experiments show that this scheme can effectively improve the accuracy of data reconstruction.When the data rate loss reaches 80%,the proposed scheme in reconstructing data is still better than the existing solution.
作者 栗风永 周刚 LI Fengyong;ZHOU Gang(School of Computer Science and Technology,Shanghai University of Electric Power,Shanghai 200090,China)
出处 《上海电力大学学报》 CAS 2020年第3期251-258,共8页 Journal of Shanghai University of Electric Power
基金 国家自然科学基金(61602295) 国家自然科学基金通用技术联合基金(U1736120)。
关键词 无线传感器网络 随机森林算法 特征选择 数据重建 wireless sensor network random forest algorithm feature selection data reconstruction
  • 相关文献

参考文献3

二级参考文献32

  • 1冯少荣.决策树算法的研究与改进[J].厦门大学学报(自然科学版),2007,46(4):496-500. 被引量:67
  • 2AKYILDIZ I F, SU W, SANKARASUBRAMANIAM Y. A survey on sensor networks [ J]. IEEE Communications Magazine, 2002, 40 (8) : 102 - 114.
  • 3DONOHO D. Compressed sensing [ J]. IEEE Transactions on Infor- mation Theory, 2006, 52(4) : 1289 - 1306.
  • 4CANDES E, ROMBERG J, TAO T. Robust uncertainty principles: Exact signal reconstruction from highly incomplete frequency infor- mation [ J]. IEEE Transactions on Information Theory, 2006, 52 (4): 489 -509.
  • 5CHEN S, DONOHO D, SAUNDERS M. Atomic decomposition by basis pursuit [ J]. SIAM Journal on Scientific Computing, 1999, 20 (1): 33-61.
  • 6KIM S, KOH K, LUSTIG M, et al. An interior-point method for large-scale 11 regularized least squares [ J]. IEEE Journal of Select- ed Topics in Signal Processing, 2007, 1(4): 606 -617.
  • 7TROPP J, GILLBERT A. Signal recovery from random measure- ments via orthogonal matching pursuit [ J]. IEEE Transactions on Information Theory, 2007, 53(12) : 4655 -4666.
  • 8DONOHO D, TSAIG Y, JEAN-LUC S. Sparse solution of under- determined linear equations by stagewise orthogonal matching pur- suit [ R]. Stanford, California, USA: Stanford University, Depart- ment of Statistics, 2006.
  • 9DAI W, MILENKOVIC O. Subspace pursuit for compressive sensing: Closing the gap between performance and complexity [ J]. IEEE Trans- actions on Information Theory, 2.009, 55(5) : 2930 - 2249.
  • 10NEEDELL D, TROPP J. CoSaMP: Iterative signal reeovery from incomplete and inaccurate samples [ J]. Applied and Computation- al Harmonic Analysis, 2009, 26(3): 301 -321.

共引文献16

同被引文献14

引证文献1

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部