Relevance vector machine technique for the inverse scattering problem

doi:10.1088/1674-1056/21/5/050204

中国物理B ›› 2012, Vol. 21 ›› Issue (5): 50204-050204.doi: 10.1088/1674-1056/21/5/050204

Relevance vector machine technique for the inverse scattering problem

王芳芳,张业荣

School of Electronic Science and Engineering, Nanjing University of Posts and Telecommunications, Nanjing 210003, China

收稿日期:2011-08-26 修回日期:2012-04-27 出版日期:2012-04-01 发布日期:2012-04-01
基金资助:
Project supported by the National Natural Science Foundation of China (Grant No. 61071022) and the Graduate Student Research and Innovation Program of Jiangsu Province, China (Grant No. CXZZ11-0381).

Relevance vector machine technique for the inverse scattering problem

Wang Fang-Fang(王芳芳) and Zhang Ye-Rong(张业荣)^†

School of Electronic Science and Engineering, Nanjing University of Posts and Telecommunications, Nanjing 210003, China

Received:2011-08-26 Revised:2012-04-27 Online:2012-04-01 Published:2012-04-01
Supported by:
Project supported by the National Natural Science Foundation of China (Grant No. 61071022) and the Graduate Student Research and Innovation Program of Jiangsu Province, China (Grant No. CXZZ11-0381).

摘要/Abstract

摘要： A novel method based on the relevance vector machine (RVM) for the inverse scattering problem is presented in this paper. The nonlinearity and the ill-posedness inherent in this problem are simultaneously considered. The nonlinearity is embodied in the relation between the scattered field and the target property, which can be obtained through the RVM training process. Besides, rather than utilizing regularization, the ill-posed nature of the inversion is naturally accounted for because the RVM can produce a probabilistic output. Simulation results reveal that the proposed RVM-based approach can provide comparative performances in terms of accuracy, convergence, robustness, generalization, and improved performance in terms of sparse property in comparison with the support vector machine (SVM) based approach.

关键词: inverse scattering problem, through-wall problem, relevance vector machine, finite-difference time-domain

Abstract: A novel method based on the relevance vector machine (RVM) for the inverse scattering problem is presented in this paper. The nonlinearity and the ill-posedness inherent in this problem are simultaneously considered. The nonlinearity is embodied in the relation between the scattered field and the target property, which can be obtained through the RVM training process. Besides, rather than utilizing regularization, the ill-posed nature of the inversion is naturally accounted for because the RVM can produce a probabilistic output. Simulation results reveal that the proposed RVM-based approach can provide comparative performances in terms of accuracy, convergence, robustness, generalization, and improved performance in terms of sparse property in comparison with the support vector machine (SVM) based approach.

Key words: inverse scattering problem, through-wall problem, relevance vector machine, finite-difference time-domain

中图分类号: (Inverse problems)

02.30.Zz

29.40.Gx (Tracking and position-sensitive detectors) 41.20.Jb (Electromagnetic wave propagation; radiowave propagation) 81.70.Ex (Nondestructive testing: electromagnetic testing, eddy-current testing)

王芳芳,张业荣. Relevance vector machine technique for the inverse scattering problem[J]. 中国物理B, 2012, 21(5): 50204-050204.

Wang Fang-Fang(王芳芳) and Zhang Ye-Rong(张业荣) . Relevance vector machine technique for the inverse scattering problem[J]. Chin. Phys. B, 2012, 21(5): 50204-050204.

参考文献 19

[1]	Ferris D D and Currie N C 1999 Proc. SPIE 3577 62
[2]	Baranoski E J 2008 J. Frank. Inst. 345 556
[3]	Soldovieri F and Solimene R 2007 IEEE Geosc. Rem. Sens. Lett. 4 513
[4]	Soldovieri F, Solimene R and Prisco G 2008 IEEE Trans. Geosci. Remot. Sens. 46 1192
[5]	Chew W C and Wang Y M 1990 IEEE Trans. Med. Imag. 9 218
[6]	Cui T J, Chew W C, Aydiner A A and Chen S Y 2001 IEEE Trans. Geosci. Remot. Sens. 39 339
[7]	Li F H, Liu Q H and Song L P 2004 IEEE Geosc. Rem. Sens. Lett. 1 107
[8]	Song L P, Yu C and Liu Q H 2005 IEEE Trans. Geosci. Remot. Sens. 43 2793
[9]	Song L P, Liu Q H, Li F H and Zhang Z Q 2005 IEEE Trans. Antennas Propagat. 53 1556
[10]	Zhang Z Q and Liu Q H 2004 IEEE Trans. Biomed. Eng. 51 544
[11]	Smola A J and Scholkopf B 2004 Stat. Comput. 14 199
[12]	Sun Z H and Jiang F 2010 Chin. Phys. B 19 110502
[13]	Meng Q F, Chen Y H and Peng Y H 2009 Chin. Phys. B 18 2194
[14]	Zhang Q H, Xiao B X and Zhu G Q 2007 Microwave Opt. Technol. Lett. 49 372
[15]	Bermani E, Boni A, Caorsi S and Massa A 2003 IEEE Trans. Geosci. Remot. Sens. 41 927
[16]	Wang F F and Zhang Y R 2011 J. Electromagn. Waves Appl. 25 75
[17]	Massa A, Boni A and Donelli M 2005 IEEE Trans. Geosci. Remot. Sens. 43 2084
[18]	Kim Y and Ling H 2009 IEEE Trans. Geosci. Remot. Sens. 47 1328
[19]	Tipping M E 2001 J. Mach. Learn. Res. 1 211

Relevance vector machine technique for the inverse scattering problem

Relevance vector machine technique for the inverse scattering problem

PDF (PC)

赞

可视化

摘要/Abstract

引用本文

使用本文

参考文献 19

相关文章 15

Metrics

本文评价

推荐阅读 0

[1]	Yun Lin(林蕴), Shuo Shen(申烁), Xiang Gao(高祥), and Liancheng Wang(汪炼成). Lattice plasmon mode excitation via near-field coupling[J]. 中国物理B, 2022, 31(1): 14214-014214.
[2]	廖文渊, 李健, 李川川, 郭小峰, 郭文涛, 刘维华, 张杨杰, 韦欣, 谭满清. Oxide-aperture-dependent output characteristics of circularly symmetric VCSEL structure[J]. 中国物理B, 2020, 29(2): 24201-024201.
[3]	周俊, 张建烁, 冼国裕, 齐琦, 顾尚志, 申承民, 成昭华, 何声太, 杨海涛. Synthesis and surface plasmon resonance of Au-ZnO Janus nanostructures[J]. 中国物理B, 2019, 28(8): 83301-083301.
[4]	陈安涛, 孙浩宇, 韩一平, 汪加洁, 崔志伟. Propagation characteristics of oblique incidence terahertz wave through non-uniform plasma[J]. 中国物理B, 2019, 28(1): 14201-014201.
[5]	王莹, 李新化. Light trapping and optical absorption enhancement in vertical semiconductor Si/SiO₂ nanowire arrays[J]. 中国物理B, 2018, 27(2): 26102-026102.
[6]	梁图禄, 邵维, 魏晓琨, 梁木生. Hybrid sub-gridding ADE-FDTD method of modeling periodic metallic nanoparticle arrays[J]. 中国物理B, 2018, 27(10): 100204-100204.
[7]	秦雪. Factorization method for inverse obstacle scattering problem in three-dimensional planar acoustic waveguides[J]. 中国物理B, 2018, 27(10): 100203-100203.
[8]	马秀荣, 徐林, 常世元, 张双根. Investigation of three-pulse photon echo in thick crystal using finite-difference time-domain method[J]. 中国物理B, 2017, 26(4): 44201-044201.
[9]	窦虎, 马红梅, 孙玉宝. Optical simulation of in-plane-switching blue phase liquid crystal display using the finite-difference time-domain method[J]. 中国物理B, 2016, 25(9): 94221-094221.
[10]	卢佳, 周怀春. Finite-difference time-domain modeling of curved material interfaces by using boundary condition equations method[J]. 中国物理B, 2016, 25(9): 90203-090203.
[11]	李旭峰, 彭伟, 赵亚丽, 王乔, 魏计林. A subwavelength metal-grating assisted sensor of Kretschmann style for investigating the sample with high refractive index[J]. 中国物理B, 2016, 25(3): 37303-037303.
[12]	魏晓琨, 邵维, 石胜兵, 张勇, 王秉中. An efficient locally one-dimensional finite-difference time-domain method based on the conformal scheme[J]. 中国物理B, 2015, 24(7): 70203-070203.
[13]	王玥, 王建国, 陈再高. Uniform stable conformal convolutional perfectly matched layer for enlarged cell technique conformal finite-difference time-domain method[J]. 中国物理B, 2015, 24(2): 24101-024101.
[14]	刘菊, 钟晓岚, 李志远. Enhanced light absorption of silicon in the near-infrared band by designed gold nanostructures[J]. 中国物理B, 2014, 23(4): 47306-047306.
[15]	刘亚文, 陈亦望, 张品, 刘宗信. A spherical higher-order finite-difference time-domain algorithm with perfectly matched layer[J]. , 2014, 23(12): 124102-124102.