基于最近邻支撑向量特征线融合算法的核爆地震识别
|
摘要
为增强核爆地震模式分类器的泛化能力以提高对核爆炸事件的准确识别能力,论文提出了一种选择支撑向量样本集来表征训练样本集的最近邻支撑向量特征线分类算法,用以训练时扩展核爆地震的训练样本库,提高分类器的泛化能力,该算法用于核爆炸和地震的识别结果发现,和最近邻特征线分类器相比,提出的算法降低了计算复杂度,但识别能力却有些许降低.对新算法的分析发现,纯粹的支撑向量集不能完全代表原始样本空间集,支撑向量比例在其中有重要作用,为发挥支撑向量比例的作用以提高核爆分类器的识别能力,提出了最近邻支撑向量特征线融合算法.最后以核爆地震数据库对上述算法进行了检验和分析,理论分析和识别结果证实,在相同的训练样本选择条件下,最近邻支撑向量特征线融合算法对于核爆炸的识别来说具有较好的泛化能力,正确识别率达到90.3%,且优于支持向量机算法和最近邻特征线算法.
Statistical pattern recognition is an important method to distinguish nuclear explosions from natural earthquakes, but because the number of nuclear explosions is few and this limits the generalization performance of the nuclear explosions' classifier, and finally affects the classification performance for the nuclear explosions. To improve the classification performance of the classifier, a nearest support vector feature line algorithm is introduced to expand the virtual nuclear explosions samples and improve the generalization performance correspondingly. However, this new algorithm can not distinguish nuclear explosions from natural earthquakes more accurately, and even is not as good as the nearest feature line algorithm. Theoretical analysis finds that the support vector set is not a good representation of the train set and the ratio of support vector plays an important role to the classifier. To improve the recognition result with the role of support vector ratio, support vector machine is integrated with nearest support vector feature line method by support vector ratio, and then a nearest support vector feature line fusion classification algorithm is presented, and experimental results and corresponding analysis show that this novel algorithm is more appropriate to the classification of underground nuclear explosions and natural earthquakes, and the accurate classification ratio is 90.3% on the data set of nuclear explosions and natural earthquakes, and moreover this classification ratio is higher than support vector machine and nearest feature line.
Statistical pattern recognition is an important method to distinguish nuclear explosions from natural earthquakes, but because the number of nuclear explosions is few and this limits the generalization performance of the nuclear explosions' classifier, and finally affects the classification performance for the nuclear explosions. To improve the classification performance of the classifier, a nearest support vector feature line algorithm is introduced to expand the virtual nuclear explosions samples and improve the generalization performance correspondingly. However, this new algorithm can not distinguish nuclear explosions from natural earthquakes more accurately, and even is not as good as the nearest feature line algorithm. Theoretical analysis finds that the support vector set is not a good representation of the train set and the ratio of support vector plays an important role to the classifier. To improve the recognition result with the role of support vector ratio, support vector machine is integrated with nearest support vector feature line method by support vector ratio, and then a nearest support vector feature line fusion classification algorithm is presented, and experimental results and corresponding analysis show that this novel algorithm is more appropriate to the classification of underground nuclear explosions and natural earthquakes, and the accurate classification ratio is 90.3% on the data set of nuclear explosions and natural earthquakes, and moreover this classification ratio is higher than support vector machine and nearest feature line.
引文
[1] 杨选辉,沈萍,刘希强等.地震与核爆识别的小波包方法.地球物理学报,2005,48(1) :148～156 Yang X H,Shen P,Liu X Q,et al.Application of method of spectral component ratio of wavelet-packets to discrimination between earthquakes and nuclear explosions.Chinese J.Geophys.(in Chinese),2005,48(1) :148～156
[2] 孙煜,范万春,许进等.基于小波包变换的地震事件分类.核电子学与探测技术,2005,25(1) :32～36 Sun Y,Fan W C,Xu J,et al.Seismic signal classification based on wavelet packet transform.Nuclear Electronics & Detection Technology(in Chinese),2005,25(1) :32～36
[3] 沈萍,郑治真.瞬态谱在地震与核爆识别中的应用.地球物理学报,1999,42(2) :231～240 Shen P,Zheng Z Z.Application of transient spectrum to discrimination of nuclear explosions and earthquakes.Chinese J.Geophys.(in Chinese),1999,42(2) :231～240
[4] 何永锋,陈晓非.利用经验格林函数识别地下核爆炸与天然地震.中国科学(D辑地球科学),2006,36(2) :177～181 He Y F,Chen X F.Using empirical Green function to discriminate underground nuclear explosions and natural earthquakes.Science in China(Ser.D Earth Sciences)(in Chinese),2006,36(2) :177～181
[5] LiuD Z,Zhao K,Zou H X,et al.Fractal analysis with applications to seismological pattern recognition of underground nuclear explosions.Signal Processing,2000,80(9) :1849～1861
[6] 刘代志,张斌,李夕海等.核爆地震模式识别中的特征相空间研究.地球物理学报,2006,49(3) :871～876 Liu D Z,Zhang B,Li X H,etal.On the feature phase space in the seismic pattern recognition of underground nuclear explosions.Chinese J.Geophys.(in Chinese),2006,49(3) :871～876
[7] 刘代志,李夕海,何志文等.基于非线性加权K-相关的核爆地震识别.核电子学与探测技术,2005,25(6) :587～589 Liu D Z,Li X H,He Z W,et al.Identification of underground nuclear explosions and natural earthquakes based on K-correlation of nonlinear weighting.Nuclear Electronics & Detection Technology(in Chinese),2005,25(6) :587～589
[8] 张斌,李夕海,苏 娟等.基于支持向量机的核爆地震自动识别.核电子学与探测技术,2005,25(1) :44～47 Zhang B,Li X H,Su J,et al.Recognition of nuclear explosion and natural earthquake based on SVM.Nuclear Electronics & Detection Technology(in Chinese),2005,25(1) :44～47
[9] 刘代志,邹红星,韦荫康等.分形分析与核爆地震模式识别.模式识别与人工智能,1997,10(2) :153～158 Liu D Z,Zou H X,Wei Y K,et al.Fractal analysis with applications to seismic pattern recognition of nuclear explosion.Pattern Recognition and Artificial Intelligence(in Chinese),1997,10(2) :153～158
[10] 刘代志,邹红星,赵克等.吸引子分析与核爆地震模式识别.电子学报,1997,25(10) :122～125 Liu D Z,Zou H X,Zhao K,et al.Attractor analysis in seismic pattern recognition of nuclear explosion.Acta Electronica Sinica(in Chinese),1997,25(10) :122～125
[11] Liu D Z,Wang R M,L I X H,et al.Seismic pattern recognition of nuclear explosion based on generalization learning algorithm of BP network and genetic algorithm.IEEE International Symposium on Neural Networks,2004． 959～964
[12] Li X H,Zhao K,Liu D Z,et al.Feature extraction and identification of underground nuclear explosion and natural earthquake based on FM~mlet transform and BP neural network. IEEE International Symposium on Neural Networks,2004． 925～930
[13] Liu D Z,Li X H,Zhang B.Feature selection and identification of underground nuclear explosion and natural earthquake based on gamma test and BP neural network.IEEE International Symposium on Neural Networks,2005． 393～398
[14] Liu G,Li X H,Liu D Z,et al.Feature extraction of underground nuclear explosions based on NMF and KNMF.IEEE International Symposium on Neural Networks,2006． 1400～1405
[15] Stan Z Li,Lu J W.Face recognition using the nearest feature line method.IEEE Trans.Neural Networks,1999,10(2) :439～443
[16] 叶伊松,武研.基于ICA和NFL分类的局部人脸识别方法.中国图象图形学报,2005,10(4) :468～472 Ye Y S,Wu Y.ICA/NFL local face recognition.Journal of Image and Graphics(in Chinese),2005,10(4) :468～472
[17] 孟继成,杨万麟.雷达目标识别中的双距离像子空间法、系统工程与电子技术,2004,26(6) :724～725 Meng J C,Yang W L.Range profile recognition of radar target based on dual range subspace.Systems Engineering and.Electronics(in Chinese),2004,26(6) :724～725
[18] 谢建平.一种基于快速最近邻特征线的汽车牌照识别方法.视频技术应用与工程.2006,4:93～96 Xie J P.A method of license character recognition based on fast nearest feature line.Application & Project of Video Technologies(in Chinese),2006,4:93～96
[19] 赵黎,祁卫,李子青等.利用该进NFL算法对镜头进行基于内容的检索.软件学报,2002,13(4) :586～590 Zhao L,Qi W,Li Z Q,et al.Content-based retrieval of video shot using the improved nearest feature line method.Journal of Software(in Chinese),2002,13(4) :586～590
[20] Nello Cristianini,John Shawe-Taylor.An Introduction to Support Vector Machines and Other Kernel-Based Learning Methods.England:Press of the University of Cambridge,2000
[21] Boser B E,Guyon I M,Vapnik V N.A training algorithm for optimal margin classifiers.In:Haussler D eds.Proceedings of the 5th Annuat ACM Workshop on Computational Learning Theory,1992． 144～152
[22] Corts C,Vapnik V N.Support Vector Networks.Machine Learning,1995,20:273～297
[23] Vapnik V N.The Nature of Statistical Learning Theory.New York:Springer-Verlag,1995
[24] Sch(?)lkopf B.Support Vector Learning.Munich:Oldenbourg Verlag,1997
[25] Vapnik V N.Statistical Learning Theory.Wiley:New York,1998
[26] Sch(?)lkopf B,Burges C J C,Smola A J.Advances in Kernel Methods-Support Vector Learning. Cambridge:MIT Press,1999
[27] 李夕海.核爆地震特征提取与识别算法研究[博士论文].西安:第二炮兵工程学院,2006 Li X H.Feature extraction and target recognition of underground nuclear explosions and natural earthquakes [Ph.D.thesis].Xi'an:The Second Artillery Institute of Engineering,2006
[2] 孙煜,范万春,许进等.基于小波包变换的地震事件分类.核电子学与探测技术,2005,25(1) :32～36 Sun Y,Fan W C,Xu J,et al.Seismic signal classification based on wavelet packet transform.Nuclear Electronics & Detection Technology(in Chinese),2005,25(1) :32～36
[3] 沈萍,郑治真.瞬态谱在地震与核爆识别中的应用.地球物理学报,1999,42(2) :231～240 Shen P,Zheng Z Z.Application of transient spectrum to discrimination of nuclear explosions and earthquakes.Chinese J.Geophys.(in Chinese),1999,42(2) :231～240
[4] 何永锋,陈晓非.利用经验格林函数识别地下核爆炸与天然地震.中国科学(D辑地球科学),2006,36(2) :177～181 He Y F,Chen X F.Using empirical Green function to discriminate underground nuclear explosions and natural earthquakes.Science in China(Ser.D Earth Sciences)(in Chinese),2006,36(2) :177～181
[5] LiuD Z,Zhao K,Zou H X,et al.Fractal analysis with applications to seismological pattern recognition of underground nuclear explosions.Signal Processing,2000,80(9) :1849～1861
[6] 刘代志,张斌,李夕海等.核爆地震模式识别中的特征相空间研究.地球物理学报,2006,49(3) :871～876 Liu D Z,Zhang B,Li X H,etal.On the feature phase space in the seismic pattern recognition of underground nuclear explosions.Chinese J.Geophys.(in Chinese),2006,49(3) :871～876
[7] 刘代志,李夕海,何志文等.基于非线性加权K-相关的核爆地震识别.核电子学与探测技术,2005,25(6) :587～589 Liu D Z,Li X H,He Z W,et al.Identification of underground nuclear explosions and natural earthquakes based on K-correlation of nonlinear weighting.Nuclear Electronics & Detection Technology(in Chinese),2005,25(6) :587～589
[8] 张斌,李夕海,苏 娟等.基于支持向量机的核爆地震自动识别.核电子学与探测技术,2005,25(1) :44～47 Zhang B,Li X H,Su J,et al.Recognition of nuclear explosion and natural earthquake based on SVM.Nuclear Electronics & Detection Technology(in Chinese),2005,25(1) :44～47
[9] 刘代志,邹红星,韦荫康等.分形分析与核爆地震模式识别.模式识别与人工智能,1997,10(2) :153～158 Liu D Z,Zou H X,Wei Y K,et al.Fractal analysis with applications to seismic pattern recognition of nuclear explosion.Pattern Recognition and Artificial Intelligence(in Chinese),1997,10(2) :153～158
[10] 刘代志,邹红星,赵克等.吸引子分析与核爆地震模式识别.电子学报,1997,25(10) :122～125 Liu D Z,Zou H X,Zhao K,et al.Attractor analysis in seismic pattern recognition of nuclear explosion.Acta Electronica Sinica(in Chinese),1997,25(10) :122～125
[11] Liu D Z,Wang R M,L I X H,et al.Seismic pattern recognition of nuclear explosion based on generalization learning algorithm of BP network and genetic algorithm.IEEE International Symposium on Neural Networks,2004． 959～964
[12] Li X H,Zhao K,Liu D Z,et al.Feature extraction and identification of underground nuclear explosion and natural earthquake based on FM~mlet transform and BP neural network. IEEE International Symposium on Neural Networks,2004． 925～930
[13] Liu D Z,Li X H,Zhang B.Feature selection and identification of underground nuclear explosion and natural earthquake based on gamma test and BP neural network.IEEE International Symposium on Neural Networks,2005． 393～398
[14] Liu G,Li X H,Liu D Z,et al.Feature extraction of underground nuclear explosions based on NMF and KNMF.IEEE International Symposium on Neural Networks,2006． 1400～1405
[15] Stan Z Li,Lu J W.Face recognition using the nearest feature line method.IEEE Trans.Neural Networks,1999,10(2) :439～443
[16] 叶伊松,武研.基于ICA和NFL分类的局部人脸识别方法.中国图象图形学报,2005,10(4) :468～472 Ye Y S,Wu Y.ICA/NFL local face recognition.Journal of Image and Graphics(in Chinese),2005,10(4) :468～472
[17] 孟继成,杨万麟.雷达目标识别中的双距离像子空间法、系统工程与电子技术,2004,26(6) :724～725 Meng J C,Yang W L.Range profile recognition of radar target based on dual range subspace.Systems Engineering and.Electronics(in Chinese),2004,26(6) :724～725
[18] 谢建平.一种基于快速最近邻特征线的汽车牌照识别方法.视频技术应用与工程.2006,4:93～96 Xie J P.A method of license character recognition based on fast nearest feature line.Application & Project of Video Technologies(in Chinese),2006,4:93～96
[19] 赵黎,祁卫,李子青等.利用该进NFL算法对镜头进行基于内容的检索.软件学报,2002,13(4) :586～590 Zhao L,Qi W,Li Z Q,et al.Content-based retrieval of video shot using the improved nearest feature line method.Journal of Software(in Chinese),2002,13(4) :586～590
[20] Nello Cristianini,John Shawe-Taylor.An Introduction to Support Vector Machines and Other Kernel-Based Learning Methods.England:Press of the University of Cambridge,2000
[21] Boser B E,Guyon I M,Vapnik V N.A training algorithm for optimal margin classifiers.In:Haussler D eds.Proceedings of the 5th Annuat ACM Workshop on Computational Learning Theory,1992． 144～152
[22] Corts C,Vapnik V N.Support Vector Networks.Machine Learning,1995,20:273～297
[23] Vapnik V N.The Nature of Statistical Learning Theory.New York:Springer-Verlag,1995
[24] Sch(?)lkopf B.Support Vector Learning.Munich:Oldenbourg Verlag,1997
[25] Vapnik V N.Statistical Learning Theory.Wiley:New York,1998
[26] Sch(?)lkopf B,Burges C J C,Smola A J.Advances in Kernel Methods-Support Vector Learning. Cambridge:MIT Press,1999
[27] 李夕海.核爆地震特征提取与识别算法研究[博士论文].西安:第二炮兵工程学院,2006 Li X H.Feature extraction and target recognition of underground nuclear explosions and natural earthquakes [Ph.D.thesis].Xi'an:The Second Artillery Institute of Engineering,2006
版权所有：© 2023 中国地质图书馆 中国地质调查局地学文献中心