Mutual-information matrix analysis for nonlinear interactions of multivariate time series

Random matrix theory (RMT) is a sophisticated technique to analyze the cross-correlations of multivariate time series, while it suffers from the limitation on characterizing the linear relationships. In this paper, we propose a new mutual-information matrix analysis to study the nonlinear interactions of multivariate time series, including: (i) The N-dimensional mutual information ranging between 0 and 1 can describe the strength of nonlinear interactions. (ii) The eigenvalues of the random mutual-information matrix yield the Marchenko–Pastur distribution, except that the dominant eigenvalue is significantly larger than the other eigenvalues. (iii) The distribution of most eigenvectors components of the random mutual-information matrix subjects to the Gaussian distribution, while the dominant eigenvector components tend to follow the uniform distribution. A large value of the N-dimensional mutual information, and the deviations from the eigenvalues distribution as well as the distribution of the eigenvectors components both imply the presence of interactions among the underlying time series. In the empirical analysis, we design a simulation which reveals the advantages of the mutual-information analysis over the RMT. We also apply the mutual-information matrix analysis to a real-world application that indicates the presence of interactions among the stock time series.

[1]  Hilmi Berk Celikoglu,et al.  Extension of Traffic Flow Pattern Dynamic Classification by a Macroscopic Model Using Multivariate Clustering , 2016, Transp. Sci..

[2]  K. Jones.  Entropy of random quantum states , 1990 .

[3]  E. Wigner,et al.  On the statistical distribution of the widths and spacings of nuclear resonance levels , 1951, Mathematical Proceedings of the Cambridge Philosophical Society.

[4]  Yi Yin,et al.  Multivariate multiscale sample entropy of traffic time series , 2016, Nonlinear Dynamics.

[5]  Sune K. Jakobsen,et al.  Mutual Information Matrices Are Not Always Positive Semidefinite , 2013, IEEE Transactions on Information Theory.

[6]  Pengjian Shang,et al.  Principal component analysis for non-stationary time series based on detrended cross-correlation analysis , 2015, Nonlinear Dynamics.

[7]  Helmut Ltkepohl,et al.  New Introduction to Multiple Time Series Analysis , 2007 .

[8]  Pengjian Shang,et al.  Multifractal Fourier detrended cross-correlation analysis of traffic signals , 2011 .

[9]  Raj Kumar Pan,et al.  Collective behavior of stock price movements in an emerging market. , 2007, Physical review. E, Statistical, nonlinear, and soft matter physics.

[10]  F. Dyson.  Statistical Theory of the Energy Levels of Complex Systems. I , 1962 .

[11]  A. Kraskov,et al.  Estimating mutual information. , 2003, Physical review. E, Statistical, nonlinear, and soft matter physics.

[12]  G. Reinsel.  Elements of Multivariate Time Series Analysis , 1995 .

[13]  Bernhard Blümich,et al.  White noise nonlinear system analysis in nuclear magnetic resonance spectroscopy , 1987 .

[14]  V. Plerou,et al.  Universal and Nonuniversal Properties of Cross Correlations in Financial Time Series , 1999, cond-mat/9902283.

[15]  M. S. Santhanam,et al.  Random matrix approach to categorical data analysis. , 2015, Physical review. E, Statistical, nonlinear, and soft matter physics.

[16]  Charles R. Johnson,et al.  Matrix analysis , 1985, Statistical Inference for Engineers and Data Scientists.

[17]  I. Jolliffe.  Principal Component Analysis , 2002 .

[18]  R. Bracewell.  The Fourier Transform and Its Applications , 1966 .

[19]  V. Plerou,et al.  Random matrix approach to cross correlations in financial data. , 2001, Physical review. E, Statistical, nonlinear, and soft matter physics.

[20]  Pengjian Shang,et al.  Measuring the uncertainty of coupling , 2015 .

[21]  Pengjian Shang,et al.  Measuring information interactions on the ordinal pattern of stock time series. , 2013, Physical review. E, Statistical, nonlinear, and soft matter physics.

[22]  Bo Wang,et al.  A method on calculating high-dimensional mutual information and its application to registration of multiple ultrasound images. , 2006, Ultrasonics.

[23]  Srinivas Peeta,et al.  Multiple measures-based chaotic time series for traffic flow prediction based on Bayesian theory , 2016, Nonlinear Dynamics.

[24]  D. Hoyer,et al.  Nonlinear analysis of heart rate and respiratory dynamics , 1997, IEEE Engineering in Medicine and Biology Magazine.

[25]  J. Hair.  Multivariate data analysis : a global perspective , 2010 .

[26]  Augusto Beléndez,et al.  Exact solution for the nonlinear pendulum , 2007 .

[27]  Schreiber,et al.  Signal separation by nonlinear projections: The fetal electrocardiogram. , 1996, Physical review. E, Statistical physics, plasmas, fluids, and related interdisciplinary topics.

[28]  Distribution of eigenvalues of detrended cross-correlation matrix , 2014 .