Disoominkfrcn between gaussian time series based on their spectral differences

Discrimination between two Gaussian time series is examined assuming that the important difference between the alternative processes is their covarianoe (spectral) structure. Using the likelihood ratio method in frequency domain a discriminant function is derived and its approximate distribution is obtained. It is demonstrated that, utilizing the Kullbadk-Leibler information measure, the frequencies or frequency bands which carry information for discrimination can be determined. Using this, it is shown that when mean functions are equal, discrimination based on the frequency with the largest discrimination information is equivalent to the classification procedure based on the best linear discriminant, Application to seismology is described by including a discussion concerning the spectral ratio discriminant for underground nuclear explosion and natural earthquake and is illustrated numerically using Rayleigh wave data from an underground and an atmospheric explosions.

[1]  Sueo Sugimoto,et al.  Spectral expressions of information measures of Gaussian time series and their relation to AIC and CAT , 1988, IEEE Trans. Inf. Theory.

[2]  T. W. Anderson,et al.  Classification into two Multivariate Normal Distributions with Different Covariance Matrices , 1962 .

[3]  George W. Housner,et al.  Generation of Artificial Earthquakes , 1964 .

[4]  Solomon Kullback,et al.  Information Theory and Statistics , 1960 .

[5]  Guy Melard,et al.  Testing for homogeneity and stability of time series , 1983 .

[6]  G. Dargahi‐Noubary.  A method for parameter estimation of non-linear regression with autocorrelated errors , 1990 .

[7]  Guy Melard,et al.  Sur un test d'galit des autocovariances de deux sries chronologiques , 1984 .

[8]  J. Alagón.  SPECTRAL DISCRIMINATION FOR TWO GROUPS OF TIME SERIES , 1989 .

[9]  D. Tjøstheim.  Autoregressive Representation of Seismic P-wave Signals with an Application to the Problem of Short-Period Discriminants , 1975 .

[10]  F. Downton,et al.  Time Series Analysis , 1961, Mathematical Gazette.

[11]  J. Sheil,et al.  The Distribution of Non‐Negative Quadratic Forms in Normal Variables , 1977 .

[12]  Arthur Gerald Brady.  Studies of response to earthquake ground motion , 1966 .

[13]  P. W. Burton,et al.  Letter to the Editors The Source-Layering Function of Underground Explosions and Earthquakes-an Application of a ' Common Path ' Method , 1971 .

[14]  D. Tjøstheim.  Some autoregressive models for short-period seismic noise , 1975, Bulletin of the Seismological Society of America.

[15]  SURFACE WAVES GENERATED BY ATMOSPHERIC NUCLEAR EXPLOSIONS. , 1970 .

[16]  A. Wood.  An F Approximation to the Distribution of a Linear Combination of Chi-squared Variables. , 1989 .

[17]  M. A. Wincek.  Applied Statistical Time Series Analysis , 1990 .

[18]  F. E. Satterthwaite.  Synthesis of variance , 1941 .

[19]  P. Molnar,et al.  Small Earthquakes and Explosions in Western North America recorded by New High Gain, Long Period Seismographs , 1969, Nature.

[20]  Yi-Ben Tsai,et al.  Amplitude spectra of surface waves from small earthquakes and underground nuclear explosions , 1971 .

[21]  G. Box.  Some Theorems on Quadratic Forms Applied in the Study of Analysis of Variance Problems, I. Effect of Inequality of Variance in the One-Way Classification , 1954 .

[22]  Jonathan D. Cryer,et al.  Time Series Analysis , 1986 .

[23]  A. W. Davis.  A Differential Equation Approach to Linear Combinations of Independent Chi-Squares , 1977 .

[24]  R. W. Farebrother,et al.  The Distribution of a Quadratic Form in Normal Variables , 1990 .

[25]  Satterthwaite Fe.  An approximate distribution of estimates of variance components. , 1946 .

[26]  P. J. Diggle,et al.  TESTS FOR COMPARING TWO ESTIMATED SPECTRAL DENSITIES , 1986 .

[27]  Peter Molnar,et al.  Excitation of seismic surface waves with periods of 15 to 70 seconds for earthquakes and underground explosions , 1971 .

[28]  P. Burton.  Estimations of Qγ-1 from Seismic Rayleigh Waves , 1974 .

[29]  David von Seggern,et al.  Theoretical and observed Rayleigh‐wave spectra for explosions and earthquakes , 1970 .