×

zbMATH — the first resource for mathematics

Discriminant analysis of time series in the presence of within-group spectral variability. (English) Zbl 1359.62374
Summary: Many studies record replicated time series epochs from different groups with the goal of using frequency domain properties to discriminate between the groups. In many applications, there exists variation in cyclical patterns from time series in the same group. Although a number of frequency domain methods for the discriminant analysis of time series have been explored, there is a dearth of models and methods that account for within-group spectral variability. This article proposes a model for groups of time series in which transfer functions are modelled as stochastic variables that can account for both between-group and within-group differences in spectra that are identified from individual replicates. An ensuing discriminant analysis of stochastic cepstra under this model is developed to obtain parsimonious measures of relative power that optimally separate groups in the presence of within-group spectral variability. The approach possesses favourable properties in classifying new observations and can be consistently estimated through a simple discriminant analysis of a finite number of estimated cepstral coefficients. Benefits in accounting for within-group spectral variability are empirically illustrated in a simulation study and through an analysis of gait variability.

MSC:
62M10 Time series, auto-correlation, regression, etc. in statistics (GARCH)
62F12 Asymptotic properties of parametric estimators
62M15 Inference from stochastic processes and spectral analysis
62H30 Classification and discrimination; cluster analysis (statistical aspects)
Software:
PhysioToolkit; sapa
PDF BibTeX XML Cite
Full Text: DOI
References:
[1] Alagón, Spectral discrimination for two groups of time series, Journal of Time Series Analysis 10 pp 203– (1989) · Zbl 04500271 · doi:10.1111/j.1467-9892.1989.tb00024.x
[2] Anderson, An Introduction to Multivariate Statistical Analysis (2003) · Zbl 1039.62044
[3] Aronszajn, Theory of reproducing kernels, American Mathematical Society Transactions 68 pp 337– (1950) · Zbl 0037.20701 · doi:10.1090/S0002-9947-1950-0051437-7
[4] Bai, Limit of the smallest eigenvalue of a large dimensional sample covariance matrix, Annals of Statistics 21 pp 1275– (1993) · Zbl 0779.60026
[5] Bloomfield, An exponential model for the spectrum of a scalar time series, Biometrika 60 pp 217– (1973) · Zbl 0261.62074 · doi:10.1093/biomet/60.2.217
[6] Bogert, Proceedings of the Symposium on Time Series Analysis pp 209– (1963)
[7] Brillinger, Time Series: Data Analysis and Theory (2001) · Zbl 0983.62056 · doi:10.1137/1.9780898719246
[8] Chatelin, The spectral approximation of linear operators with applications to computation of eigenelements of differential and integral operators., SIAM Review 23 pp 495– (1981) · Zbl 0472.65048 · doi:10.1137/1023099
[9] Cheng, Asymptotic probabilities of misclassification of two discriminant functions in cases of high dimensional data, Statistics & Probability Letters 67 pp 9– (2004) · Zbl 1081.62042 · doi:10.1016/j.spl.2003.12.001
[10] Dargahi-Noubary, Spectral ratio discriminants and information theory, Journal of Time Series Analysis 16 pp 201– (1981) · Zbl 0509.62052 · doi:10.1111/j.1467-9892.1995.tb00230.x
[11] Delaigle, Achieving near perfect classification for functional data, Journal of the Royal Statistical Society, Series B 74 pp 267– (2012) · doi:10.1111/j.1467-9868.2011.01003.x
[12] Diggle, Spectral analysis of replicated biomedical time series, Applied Statistics 46 pp 37– (1997) · Zbl 0893.62108
[13] Eubank, Canonical correlation for stochastic processes, Stochastic Processes and their Applications 118 pp 1634– (2008) · Zbl 1145.62048 · doi:10.1016/j.spa.2007.10.006
[14] Freyermuth, Tree-structured wavelet estimation in a mixed effects model for spectra of replicated time series, Journal of the American Statistical Association 105 pp 634– (2010) · Zbl 1392.62282 · doi:10.1198/jasa.2010.tm09132
[15] Goldberger, PhysioBank, PhysioToolkit, and PhysioNet: components of a new research resource for complex physiologic signals, Circulation 101 pp e215– (2000) · doi:10.1161/01.CIR.101.23.e215
[16] Hausdorff, Gait variability: methods, modeling, and meaning, Journal of Neuroengineering and Rehabilitation 2 (2005) · doi:10.1186/1743-0003-2-19
[17] Hausdorff, Dynamic markers of altered gait rhythm in amyotrophic lateral sclerosis, Journal of Applied Physiology 88 pp 2045– (2000)
[18] Henmi, Spectral analysis of gait variability of stride interval time series: comparision of young, elderly and Parkinson’s disease patients, Journal of Physical Therapy Science 21 pp 105– (2009) · doi:10.1589/jpts.21.105
[19] Iannaccone, Semiparametric models and inference for biomedical time series with extra-variation, Biostatistics 2 pp 261– (2001) · Zbl 1097.62561 · doi:10.1093/biostatistics/2.3.261
[20] Janas, Consistency for non-linear functionals of the periodogram of tapered data, Journal of Time Series Analysis 16 pp 585– (1995) · Zbl 0852.62088 · doi:10.1111/j.1467-9892.1995.tb00257.x
[21] Johnson, Applied Multivariate Statistical Analysis (2007)
[22] Kakizawa, Discrimination and clustering for multivariate time series, Joural of the Americal Statistical Association 93 pp 328– (1998) · doi:10.1080/01621459.1998.10474114
[23] Krafty, Penalized multivariate Whittle likelihood for power spectrum estimation, Biometrika 100 pp 447– (2013) · Zbl 1284.62581 · doi:10.1093/biomet/ass088
[24] Krafty, Functional mixed effects spectral analysis, Biometrika 98 pp 583– (2011) · Zbl 1231.62168 · doi:10.1093/biomet/asr032
[25] Ombao, A simple generalised crossvalidation method of span selection for periodogram smoothing, Biometrika 88 pp 1186– (2001) · Zbl 1008.62092 · doi:10.1093/biomet/88.4.1186
[26] O’Neill, Error rates of non-Bayes classification rules and the robustness of Fisher’s linear discriminant function, Biometrika 79 pp 177– (1992) · Zbl 0754.62047 · doi:10.1093/biomet/79.1.177
[27] Parzen, An approach to time series analysis, Annals of Mathematical Statistics 32 pp 951– (1961) · Zbl 0107.13801 · doi:10.1214/aoms/1177704840
[28] Parzen, Extraction and detection problems and reproducing kernel Hilbert spaces, Journal of SIAM Control Series A 1 pp 35– (1962) · Zbl 0199.21904
[29] Percival, Spectral Analysis for Physical Applications: Multitaper and Conventional Univariate Techniques (1993) · Zbl 0796.62077 · doi:10.1017/CBO9780511622762
[30] Saavedra, Spectral analysis with replicated time series, Communications in Statistics - Theory and Methods 29 pp 2343– (2000) · Zbl 1019.62088 · doi:10.1080/03610920008832610
[31] Saavedra, Estimation of population spectrum for linear processes with random coefficients, Computational Statistics 23 pp 79– (2008) · doi:10.1007/s00180-007-0069-5
[32] Shin, An extension of Fisher’s discriminant analysis for stochastic processes, Journal of Multivariate Analysis 99 pp 1191– (2008) · Zbl 1141.62053 · doi:10.1016/j.jmva.2007.08.001
[33] Shumway, Classification, Pattern Recognition and Reduction of Dimensionality, Handbook of Statistics pp 1– (1982) · doi:10.1016/S0169-7161(82)02004-5
[34] Shumway, Time Series Analysis and its Applications: With R Examples, 3. ed. (2011) · Zbl 1276.62054 · doi:10.1007/978-1-4419-7865-3
[35] Shumway, Linear discriminant functions for stationary time series, Journal of the American Statistical Association 69 pp 948– (1974) · Zbl 0296.62085 · doi:10.1080/01621459.1974.10480234
[36] Thomson, Spectrum estimation and harmonic analysis, Proceedings of the IEEE 70 pp 1055– (1982) · doi:10.1109/PROC.1982.12433
[37] Zhang, Discriminant analysis for stationary vector time series, Journal of Time Series Analysis 15 pp 117– (1994) · Zbl 0794.62064 · doi:10.1111/j.1467-9892.1994.tb00180.x
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. It attempts to reflect the references listed in the original paper as accurately as possible without claiming the completeness or perfect precision of the matching.