Testing equality of spectral density operators for functional processes

The problem of comparing the entire second order structure of two functional processes is considered and a L2-type statistic for testing equality of the corresponding spectral density operators is investigated. The test statistic evaluates, over all frequencies, the Hilbert-Schmidt distance between the two estimated spectral density operators. Under certain assumptions, the limiting distribution under the null hypothesis is derived. A novel frequency domain bootstrap method is introduced, which leads to a more accurate approximation of the distribution of the test statistic under the null than the large sample Gaussian approximation derived. Under quite general conditions, asymptotic validity of the bootstrap procedure is established for estimating the distribution of the test statistic under the null. Furthermore, consistency of the bootstrap-based test under the alternative is proved. Numerical simulations show that, even for small samples, the bootstrapbased test has a very good size and power behavior. An application to a bivariate real-life functional time series illustrates the methodology proposed.

[1]  P. Alam ‘T’ , 2021, Composites Engineering: An A–Z Guide.

[2]  Ron Reeder,et al.  Estimation of the mean of functional time series and a two‐sample problem , 2011, 1105.0019.

[3]  Gorjan Alagic,et al.  #p , 2019, Quantum information & computation.

[4]  J. Franke,et al.  A residual-based bootstrap for functional autoregressions , 2019, 1905.07635.

[5]  T. Sapatinas,et al.  Testing equality of spectral density operators for functional linear processes , 2018, 1804.03366.

[6]  Ricardo Fraiman,et al.  Editorial for the Special Issue on Functional Data Analysis and Related Topics , 2019, J. Multivar. Anal..

[7]  Efstathios Paparoditis,et al.  Spectral Density Based Goodness-of-Fit Tests for Time Series Models , 2000 .

[8]  Holger Dette,et al.  Bootstrapping frequency domain tests in multivariate time series with an application to comparing spectral densities , 2009 .

[9]  P. Alam,et al.  R , 1823, The Herodotus Encyclopedia.

[10]  P. Alam ‘E’ , 2021, Composites Engineering: An A–Z Guide.

[11]  Efstathios Paparoditis Sieve bootstrap for functional time series , 2018, The Annals of Statistics.

[12]  Clifford M. Hurvich,et al.  Data-Driven Choice of a Spectrum Estimate: Extending the Applicability of Cross-Validation Methods , 1985 .

[13]  T. Sapatinas,et al.  Moving block and tapered block bootstrap for functional time series with an application to the $K$-sample mean problem , 2017, Bernoulli.

[14]  Alois Kneip,et al.  Common Functional Principal Components , 2006, 0901.4252.

[15]  Peter Bloomfield,et al.  DETERMINING THE BANDWIDTH OF A KERNEL SPECTRUM ESTIMATE , 1987 .

[16]  Piotr Kokoszka,et al.  Special issue on functional data analysis , 2017 .

[17]  M. Hallin,et al.  Dynamic functional principal components , 2015 .

[18]  M. Eichler,et al.  Locally Stationary Functional Time Series , 2016, 1602.05125.

[19]  Richard A. Davis,et al.  Time Series: Theory and Methods , 2013 .

[20]  柿沢 佳秀,et al.  Asymptotic theory of statistical inference for time series , 2000 .

[21]  Han Lin Shang,et al.  Bootstrap methods for stationary functional time series , 2016, Stat. Comput..

[22]  Heng Peng,et al.  Two Samples Tests for Functional Data , 2010 .

[23]  Victor M. Panaretos,et al.  Fourier analysis of stationary time series in function space , 2013, 1305.2073.

[24]  Piotr Kokoszka,et al.  Testing the Equality of Covariance Operators in Functional Samples , 2011, 1104.4049.

[25]  Uniform Convergence of Multivariate Spectral Density Estimates , 2015, 1505.03659.

[26]  Piotr Kokoszka,et al.  Inference for Functional Data with Applications , 2012 .

[27]  Xiaofeng Shao,et al.  Two sample inference for the second-order property of temporally dependent functional data , 2015, 1506.00847.

[28]  Herold Dehling,et al.  Bootstrap for dependent Hilbert space-valued random variables with application to von Mises statistics , 2015, J. Multivar. Anal..

[29]  Victor M. Panaretos,et al.  Detecting and Localizing Differences in Functional Time Series Dynamics: A Case Study in Molecular Biophysics , 2016 .

[30]  Juan Vilar,et al.  Detection of outliers in functional time series , 2015 .

[31]  John H. Maddocks,et al.  Second-Order Comparison of Gaussian Random Functions and the Geometry of DNA Minicircles , 2010 .

[32]  Wolfgang Härdle,et al.  On Bootstrapping Kernel Spectral Estimates , 1992 .

[33]  Peter F. de Jong,et al.  A central limit theorem for generalized quadratic forms , 1987 .

[34]  Holger Dette,et al.  Pivotal tests for relevant differences in the second order dynamics of functional time series , 2020 .

[35]  P. Robinson,et al.  AUTOMATIC FREQUENCY DOMAIN INFERENCE ON SEMIPARAMETRIC AND NONPARAMETRIC MODELS , 1991 .

[36]  E. Mammen,et al.  Comparing Nonparametric Versus Parametric Regression Fits , 1993 .

[37]  Michael Eichler,et al.  Testing nonparametric and semiparametric hypotheses in vector stationary processes , 2008 .

[38]  Frédéric Ferraty,et al.  Kernel Regression Estimation for Functional Data , 2018, Oxford Handbooks Online.

[39]  Piercesare Secchi,et al.  Distances and inference for covariance operators , 2014 .

[40]  Clément Cerovecki,et al.  On the CLT for discrete Fourier transforms of functional time series , 2015, J. Multivar. Anal..

[41]  Theofanis Sapatinas,et al.  Bootstrap-based testing of equality of mean functions or equality of covariance operators for functional data , 2016 .