基于反馈神经网络的稀疏信号恢复的优化算法

doi:10.11772/j.issn.1001-9081.2017.09.2590

计算机应用 ›› 2017, Vol. 37 ›› Issue (9): 2590-2594.DOI: 10.11772/j.issn.1001-9081.2017.09.2590

基于反馈神经网络的稀疏信号恢复的优化算法

汪星星, 李国成

北京信息科技大学理学院, 北京 100192

收稿日期:2017-03-23 修回日期:2017-05-31 出版日期:2017-09-10 发布日期:2017-09-13
通讯作者: 汪星星,wangxx501@163.com
作者简介:汪星星(1991-),女,湖北黄冈人,硕士研究生,主要研究方向:神经网络优化计算;李国成(1964-),男,河北承德人,教授,博士,主要研究方向:神经网络优化计算。
基金资助:
国家自然科学基金资助项目（61473325）。

Sparse signal reconstruction optimization algorithm based on recurrent neural network

WANG Xingxing, LI Guocheng

College of Science, Beijing Information Science and Technology University, Beijing 100192, China

Received:2017-03-23 Revised:2017-05-31 Online:2017-09-10 Published:2017-09-13
Supported by:
This work is partially supported by the National Natural Science Foundation of China (61473325).

摘要/Abstract

摘要： 针对稀疏信号的重构问题，提出了一种基于反馈神经网络（RNN）的优化算法。首先，需要对信号进行稀疏表示，将数学模型化为优化问题；接着，基于l₀范数是非凸且不可微的函数，并且该优化问题是NP难的，因此在测量矩阵A满足有限等距性质（RIP）的前提下，提出等价优化问题；最后，通过建立相应的Hopfield反馈神经网络模型来解决等价的优化问题，从而实现稀疏信号的重构。实验结果表明，在不同观测次数m下，对比RNN算法和其他三种算法的相对误差，发现RNN算法相对误差小，且需要的观测数也少，能够高效地重构稀疏信号。

关键词: l₀最优化, 反馈神经网络, 有限等距性, 能量函数

Abstract: Aiming at the problem of sparse signal reconstruction, an optimization algorithm based on Recurrent Neural Network (RNN) was proposed. Firstly, the signal sparseness was represented, and the mathematical model was transformed into an optimization problem. Then, based on the fact that the l₀-norm is a non-convex and non-differentiable function, and the optimization problem is NP-hard, under the premise that the measurement matrix A met Restricted Isometry Property (RIP), the equivalent optimization problem was proposed. Finally, the corresponding Hopfield RNN model was established to solve the equivalent optimization problem, so as to reconstruct sparse signals. The experimental results show that under different observation number m, compared the RNN algorithm and the other three algorithms, it is found that the relative error of the RNN algorithm is smaller and the observations number is smaller, and the RNN algorithm can reconstruct the sparse signals efficiently.

Key words: l₀ optimization, Recurrent Neural Network (RNN), Restricted Isometry Property (RIP), energy function

中图分类号:

TP18

汪星星, 李国成. 基于反馈神经网络的稀疏信号恢复的优化算法[J]. 计算机应用, 2017, 37(9): 2590-2594.

WANG Xingxing, LI Guocheng. Sparse signal reconstruction optimization algorithm based on recurrent neural network[J]. Journal of Computer Applications, 2017, 37(9): 2590-2594.

参考文献

[1] CANDÈS E J, ROMBERG J K, TAO T. Stable signal recovery from incomplete and inaccurate measurements[J]. Communications on Pure and Applied Mathematics, 2006, 59(8):1207-1223.
[2] CANDÈS E J, ROMBERG J, TAO T. Robust uncertainty principles:exact signal reconstruction from highly incomplete frequency information[J]. IEEE Transactions on Information Theory, 2006, 52(2):489-509.
[3] 尹宏鹏,刘兆栋,柴毅,等.压缩感知综述[J].控制与决策,2013,28(10):1441-1445.(YIN H P, LIU Z D, CHAI Y, et al. Survey of compressed sensing[J]. Control and Decision, 2013, 28(10):1441-1445.)
[4] FOUCART S, LAI M J. Sparsest solutions of underdetermined linear systems via e_q-minimization for 0<q ≤ 1[J]. Applied and Computational Harmonic Analysis, 2009, 26(3):395-407.
[5] BLUMENSATH T, DAVIES M E. Gradient pursuits[J]. IEEE Transactions on Signal Processing, 2008, 56(6):2370-2382.
[6] DAI W, MILENKOVIC O. Subspace pursuit for compressive sensing signal reconstruction[J]. IEEE Transactions on Information Theory, 2009, 55(5):2230-2249.
[7] MALLAT S G, ZHANG Z F. Matching pursuits with time-frequency dictionaries[J]. IEEE Transactions on Signal Processing, 1993, 41(12):3397-3415.
[8] FIGUEIREDO M A T, NOWAK R D, WRIGHT S J. Gradient projection for sparse reconstruction:application to compressed sensing and other inverse problems[J]. IEEE Journal of Selected Topics in Signal Processing, 2007, 1(4):586-597.
[9] CHEN S S, DONOHO D L, SAUNDERS M A. Atomic decomposition by basis pursuit[J]. SIAM Journal on Scientific Computing, 1998, 20(1):33-61.
[10] NEEDELL D, TROPP J A. CaSaMP:iterative signal recovery from incomplete and inaccurate samples[J]. Applied and Computational Harmonic Analysis, 2008, 26(3):301-321.
[11] 李珅,马彩文,李艳,等.压缩感知重构算法综述[J].红外与激光工程,2013,42(S1):225-232.(LI S, MA C W, LI Y, et al. Survey on reconstruction algorithm based on compressive sensing[J]. Infrared and Laser Engineering, 2013, 42(S1):225-232.)
[12] BOGACKI P, SHAMPINE L F. A 3(2) pair of Runge-Kutta formulas[J]. Applied Mathematics Letters, 1989, 2(4):321-325.
[13] 李国成,宋士吉,吴澄.解决非可微凸优化问题的次梯度反馈神经网络[J].中国科学(E辑),2006,36(8):811-824.(LI G C, SONG S J, WU C. Sub-gradient feedback neural network for solving non-differentiable convex optimization[J]. SCIENCE CHINA Sev. E Information Sciences, 2006, 36(8):811-824.)
[14] 曾喆昭.神经网络优化方法及其在信息处理中的应用研究[D].长沙:湖南大学,2008:62-73.(ZENG Z Z. Neural network optimization method and its application in information processing[D]. Changsha:Hunan University, 2008:62-73.)
[15] XIA Y, FENG G, WANG J. A novel recurrent neural network for solving nonlinear optimization problems with inequality constraints[J]. IEEE Transactions on Neural Networks, 2008, 19(8):1340-1353.
[16] LIU Q, WANG J. A one-layer recurrent neural network with a discontinuous hard-limiting activation function for quadratic programming[J]. IEEE Transactions on Neural Networks, 2008, 19(4):558-570.
[17] LEUNG C S, SUM J, CONSTANTINIDES A G. Recurrent networks for compressive sampling[J]. Neurocomputing, 2014, 129:298-305.
[18] LIU Q, WANG J. A one-layer recurrent neural network with a discontinuous activation function for linear programming[J]. Neural Computation, 2008, 20(5):1366-1383.

基于反馈神经网络的稀疏信号恢复的优化算法

Sparse signal reconstruction optimization algorithm based on recurrent neural network

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

编辑推荐

Metrics

[1]	焦海林, 郭玉英, 朱正为. 基于加速度补偿的无人机吊挂飞行抗摆控制[J]. 计算机应用, 2021, 41(2): 604-610.
[2]	于永斌, 戚敏惠, 尼玛扎西, 王琳. 基于阈值自适应忆阻器Hopfield神经网络的关联规则挖掘算法[J]. 计算机应用, 2019, 39(3): 728-733.
[3]	蔡兴泉, 孙辰, 葛亚坤. 限制失真的网格参数化方法[J]. 计算机应用, 2019, 39(10): 3034-3039.
[4]	丁飞飞, 杨文元. 信息熵约束下的视频目标分割[J]. 计算机应用, 2018, 38(10): 2782-2787.
[5]	王强, 张培林, 王怀光, 杨望灿, 陈彦龙. 压缩感知中测量矩阵构造综述[J]. 计算机应用, 2017, 37(1): 188-196.
[6]	王雷, 黄晨雪. 改进的分层马尔可夫随机场彩色图像分割算法[J]. 计算机应用, 2016, 36(9): 2576-2579.
[7]	李明旭, 邓欣, 王进, 王潇, 张笑谋. 基于人工神经网络的秀丽隐杆线虫趋温性行为的建模与仿真[J]. 计算机应用, 2016, 36(7): 1909-1913.
[8]	范海菊, 刘国奇. 基于核自组织映射的有监督主动轮廓图像分割[J]. 计算机应用, 2016, 36(10): 2832-2836.
[9]	郑坚周尚波. 基于神经网络的电影票房预测建模[J]. 计算机应用, 2014, 34(3): 742-748.
[10]	张丽红何树成. 基于差值绝对值之和和置信传播的快速收敛立体匹配算法[J]. 计算机应用, 2014, 34(3): 824-827.
[11]	樊博杨晓梅胡学姝. 基于压缩感知的超分辨率图像重建 [J]. 计算机应用, 2013, 33(02): 480-483.
[12]	郑伯川桑永胜. 求解最短路径问题的Lotka-Volterra回复式神经网络模型[J]. 计算机应用, 2012, 32(07): 1965-1968.
[13]	杨斌范媛媛王继东. 点云模型上近似测地线的计算[J]. 计算机应用, 2011, 31(04): 1050-1052.
[14]	蒲静傅饶胡栋卢璇. 基于动态特征的突发暴力事件辨识算法[J]. 计算机应用, 2011, 31(02): 410-412.
[15]	蒲静胡栋. 基于时空Markov随机场的人体异常行为识别算法[J]. 计算机应用, 2010, 30(8): 2238-2240.