基于Parzen窗的高阶统计量特征降维方法
|
摘要
高阶统计量通常能比低阶统计量提取更多原数据的信息,但是较高的阶数带来了较高的时间复杂度.基于Parzen窗估计构造了高阶统计量,通过论证得出:对于所提出的核协方差成分分析(KCCA)方法,通过调节二阶统计量广义D-vs-E的参数就能够达到整合高阶统计量的目的,而无需计算更高阶统计量.即核协方差成分分析方法能够对高阶统计量的特征降维的同时,又不增加计算复杂性.
The high-order statistics method can often extract more information regarding original data than a low-order statistics;yet in the meantime create higher time complexity.The high-order statistics methods were constructed by utilizing estimation based on Parzen window.It was revealed that the kernel covariance component analysis(KCCA) method proposed earlier by the researchers,contained useful information on the high-order statistics and could be obtained by only adjusting the parameters of the proposed generalized D-vs-E.Also based on the second order statistics,the heavy computational burden about the high-order statistics can be avoided.That is to say,the KCCA method can accomplish the feature reduction of high-order statistics without increasing its computational complexity.
The high-order statistics method can often extract more information regarding original data than a low-order statistics;yet in the meantime create higher time complexity.The high-order statistics methods were constructed by utilizing estimation based on Parzen window.It was revealed that the kernel covariance component analysis(KCCA) method proposed earlier by the researchers,contained useful information on the high-order statistics and could be obtained by only adjusting the parameters of the proposed generalized D-vs-E.Also based on the second order statistics,the heavy computational burden about the high-order statistics can be avoided.That is to say,the KCCA method can accomplish the feature reduction of high-order statistics without increasing its computational complexity.
引文
[1]LU Wei,SUN Wei,CHUNG Fulai,et al.Revealing digitalfakery using multiresolution decomposition and higher orderstatistics[J].Engineering Applications of Artificial Intelli-gence,2011,24(4):666-672.
[2]MUNSHI D,KITCHING T,HEAVENS A,et al.Higherorder statistics for three-dimensional shear and flexion[J].Monthly Notices of the Royal Astronomical Society,2011,416(3):1629-1653.
[3]NAPOLITANO A,TESAURO M.Almost-periodic higherorder statistic estimation[J].IEEE Transactions on Infor-mation Theory,2011,57(1):514-533.
[4]KALIDINDI S,NIEZGODA S,SALEM A.Microstructureinformatics using higher-order statistics and efficient data-mining protocols[J].JOM,2011,63(4):34-41.
[5]AGUERA-PEREZ A,PALOMARES-SALAS J,De LA RO-SA J,et al.Characterization of electrical sags and swells u-sing higher-order statistical estimators[J].Measurement,2011,44(8):1453-1460.
[6]LABBI A,BOSCH H,PELLEGRINI C.High order statis-tics for image classification[J].International Journal ofNeural Systems,2001,11(4):371-378.
[7]张丽琴,詹麒,朱培民,等.地震散射波场的高阶统计分析[J].石油地球物理勘探,2004,39(1):45-49.ZHANG Liqin,ZHAN Qi,ZHU Peimin,et al.High-orderstatistic analysis of seismic dispersive wavefield[J].Oil Ge-ophysical Prospecting,2004,39(1):45-49.
[8]ANASTASSIOU G,DUMAN O.High order statistical fuzzyKorovkin theory[J].Stochastic Analysis and Applications,2009,27(3):543-554.
[9]NIKORA V,GORING D.Martian topography:scaling,craters,and high-order statistics[J].Mathematical Geolo-gy,2005,37(4):337-355.
[10]PORAT B,FRIEDLANDER B.Direction finding algo-rithms based on high-order statistics[J].IEEE Transac-tions on Signal Processing,1991,39(9):2016-2024.
[11]COURNAPEAU D,KAWAHARA T.Voice activity detec-tion based on high order statistics and online EM algorithm[J].IEICE Transactions on Information and Systems,2008,91(12):2854-2861.
[12]REN Hsuan,DU Qian,WANG Jing,et al.Automatic tar-get recognition for hyperspectral imagery using high-orderstatistics[J].IEEE Transactions on Aerospace and Elec-tronic Systems,2006,42(4):1372-1385.
[13]TAOUFIKI M,ADIB A,ABOUTAJDINE D.Blind separa-tion of any source distributions via high-order statistics[J].Signal Processing,2007,87(8):1882-1889.
[14]YUAN Jinghe,HU Ziqiang.High-order statistical blinddeconvolution of spectroscopic data with a Gauss-Newtonalgorithm[J].Applied Spectroscopy,2006,60(6):692-697.
[15]JENSSEN R.Kernel entropy component analysis[J].IEEE Transactions on Pattern Analysis and Machine Intel-ligence,2010,32(5):847-860.
[16]闫晓波,王士同,郭慧玲.核协方差成分分析方法及其在聚类中的应用[J].计算机科学,2012,39(9):229-234.YAN Xiaobo,WANG Shitong,GUO Huiling.Kernel co-variance component analysis and its application in cluste-ring[J].Computer Science,2012,39(9):229-234.
[17]RENYI A.On measures of entropy and information[C]//Proceedings of the 4th Berkeley Symposium on MathematicalStatistics and Probability.Berkeley,USA,1961:547-561.
[18]PARZEN E.On estimation of a probability density functionand mode[J].The Annals of Mathematical Statistics,1962,33(3):1065-1076.
[19]DENG Zhaohong,CHUNG Fulai,WANG Shitong.FRS-DE:fast reduced set density estimator using minimal en-closing ball approximation[J].Pattern Recognition,2008,41(4):1363-1372.
[20]KOLLIOS G,GUNOPULOS D,KOUDAS N,et al.Effi-cient biased sampling for approximate clustering and outlierdetection in large data sets[J].IEEE Transactions onKnowledge and Date Engineering,2003,15(5):1170-1187.
[2]MUNSHI D,KITCHING T,HEAVENS A,et al.Higherorder statistics for three-dimensional shear and flexion[J].Monthly Notices of the Royal Astronomical Society,2011,416(3):1629-1653.
[3]NAPOLITANO A,TESAURO M.Almost-periodic higherorder statistic estimation[J].IEEE Transactions on Infor-mation Theory,2011,57(1):514-533.
[4]KALIDINDI S,NIEZGODA S,SALEM A.Microstructureinformatics using higher-order statistics and efficient data-mining protocols[J].JOM,2011,63(4):34-41.
[5]AGUERA-PEREZ A,PALOMARES-SALAS J,De LA RO-SA J,et al.Characterization of electrical sags and swells u-sing higher-order statistical estimators[J].Measurement,2011,44(8):1453-1460.
[6]LABBI A,BOSCH H,PELLEGRINI C.High order statis-tics for image classification[J].International Journal ofNeural Systems,2001,11(4):371-378.
[7]张丽琴,詹麒,朱培民,等.地震散射波场的高阶统计分析[J].石油地球物理勘探,2004,39(1):45-49.ZHANG Liqin,ZHAN Qi,ZHU Peimin,et al.High-orderstatistic analysis of seismic dispersive wavefield[J].Oil Ge-ophysical Prospecting,2004,39(1):45-49.
[8]ANASTASSIOU G,DUMAN O.High order statistical fuzzyKorovkin theory[J].Stochastic Analysis and Applications,2009,27(3):543-554.
[9]NIKORA V,GORING D.Martian topography:scaling,craters,and high-order statistics[J].Mathematical Geolo-gy,2005,37(4):337-355.
[10]PORAT B,FRIEDLANDER B.Direction finding algo-rithms based on high-order statistics[J].IEEE Transac-tions on Signal Processing,1991,39(9):2016-2024.
[11]COURNAPEAU D,KAWAHARA T.Voice activity detec-tion based on high order statistics and online EM algorithm[J].IEICE Transactions on Information and Systems,2008,91(12):2854-2861.
[12]REN Hsuan,DU Qian,WANG Jing,et al.Automatic tar-get recognition for hyperspectral imagery using high-orderstatistics[J].IEEE Transactions on Aerospace and Elec-tronic Systems,2006,42(4):1372-1385.
[13]TAOUFIKI M,ADIB A,ABOUTAJDINE D.Blind separa-tion of any source distributions via high-order statistics[J].Signal Processing,2007,87(8):1882-1889.
[14]YUAN Jinghe,HU Ziqiang.High-order statistical blinddeconvolution of spectroscopic data with a Gauss-Newtonalgorithm[J].Applied Spectroscopy,2006,60(6):692-697.
[15]JENSSEN R.Kernel entropy component analysis[J].IEEE Transactions on Pattern Analysis and Machine Intel-ligence,2010,32(5):847-860.
[16]闫晓波,王士同,郭慧玲.核协方差成分分析方法及其在聚类中的应用[J].计算机科学,2012,39(9):229-234.YAN Xiaobo,WANG Shitong,GUO Huiling.Kernel co-variance component analysis and its application in cluste-ring[J].Computer Science,2012,39(9):229-234.
[17]RENYI A.On measures of entropy and information[C]//Proceedings of the 4th Berkeley Symposium on MathematicalStatistics and Probability.Berkeley,USA,1961:547-561.
[18]PARZEN E.On estimation of a probability density functionand mode[J].The Annals of Mathematical Statistics,1962,33(3):1065-1076.
[19]DENG Zhaohong,CHUNG Fulai,WANG Shitong.FRS-DE:fast reduced set density estimator using minimal en-closing ball approximation[J].Pattern Recognition,2008,41(4):1363-1372.
[20]KOLLIOS G,GUNOPULOS D,KOUDAS N,et al.Effi-cient biased sampling for approximate clustering and outlierdetection in large data sets[J].IEEE Transactions onKnowledge and Date Engineering,2003,15(5):1170-1187.
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