Abstract
We consider paired and two-sample break-detection procedures for vectorial observations and multivariate time series. The new methods involve L2-type criteria based on empirical characteristic functions and are easy to compute regardless of dimension. We obtain asymptotic results that allow for application of the methods to a wide range of settings involving on-line as well as retrospective circumstances with dependence between the two time series as well as with dependence within each series. In the ensuing Monte Carlo study the new detection methods are implemented by means of resampling procedures which are properly adapted to the type of data at hand, be it independent or paired, autoregressive or GARCH structured, medium or heavy-tailed. The new methods are also applied on a real dataset from the financial sector over a time period which includes the Brexit referendum.
Similar content being viewed by others
References
Alippi C, Boracchi G, Carrera D, Roveri M (2015) Change detection in multivariate datastreams: likelihood and detectability loss. In: Proceedings of the twenty-fifth international joint conference on artificial intelligence, pp 1368–1374
Aminikhanghahi S, Cook DJ (2017) A survey of methods for time series change point detection. Knowl Inf Syst 51(2):339–367
Aue A, Hörmann S, Horváth L, Reimherr M (2009) Break detection in the covariance structure of multivariate time series models. Ann Stat 37(6B):4046–4087
Billingsley P (2013) Convergence of probability measures. Wiley, Hoboken
Chan NH, Yau CY, Zhang RM (2014) Group LASSO for structural break time series. J Am Sta Assoc 109(506):590–599
Chen F, Meintanis SG, Zhu L (2019) On some characterizations and multidimensional criteria for testing homogeneity, symmetry and independence. J Multivar Anal 173:125–144
Doukhan P (1994) Mixing: properties and examples, vol 85. Lecture notes in statistics. Springer, New York
Doukhan P, Lang G, Leucht A, Neumann MH (2015) Dependent wild bootstrap for the empirical process. J Time Ser Anal 36(3):290–314
Edelmann D, Fokianos K, Pitsillou M (2019) An updated literature review of distance correlation and its applications to time series. In Stat Rev 87(2):237–262
Galeano P, Peña D (2007) Covariance changes detection in multivariate time series. J Stat Plan Inference 137(1):194–211
Groen JJ, Kapetanios G, Price S (2013) Multivariate methods for monitoring structural change. J App Econ 28(2):250–274
Henze N, Klar B, Zhu LX et al (2005) Checking the adequacy of the multivariate semiparametric location shift model. J Multivar Anal 93(2):238–256
Hlávka Z, Hušková M, Kirch C, Meintanis SG (2012) Monitoring changes in the error distribution of autoregressive models based on Fourier methods. Test 21(4):605–634
Hlávka Z, Hušková M, Kirch C, Meintanis SG (2017a) Fourier-type tests involving martingale difference processes. Econ Rev 36(4):468–492
Hlávka Z, Hušková M, Meintanis SG (2017b) Change point detection with multivariate observations based on characteristic functions. In: Ferger D, González Manteiga W, Schmidt T, Wang JL (eds) From statistics to mathematical finance. Festschrift in Honour of Winfried Stute. Springer, New York, pp 273–290
Holmes M, Kojadinovic I, Quessy JF (2013) Nonparametric tests for change-point detection à la Gombay and Horváth. J Multivar Anal 115:16–32
Horváth L, Rice G (2014) Extensions of some classical methods in change point analysis. Test 23(2):219–255
Horváth L, Hušková M, Kokoszka P, Steinebach J (2004) Monitoring changes in linear models. J Stat Plan Inference 126(1):225–251
Hušková M, Kirch C (2012) Bootstrapping sequential change-point tests for linear regression. Metrika 75(5):673–708
Hušková M, Meintanis SG (2006) Change point analysis based on empirical characteristic functions. Metrika 63(2):145–168
Hušková M, Meintanis SG (2008) Tests for the multivariate k-sample problem based on the empirical characteristic function. J Nonparam Stat 20(3):263–277
Ibragimov I, Chasminskij R (1981) Statistical estimation: asymptotic theory. Springer, New York
James NA, Matteson DS (2014) ecp: an R package for nonparametric multiple change point analysis of multivariate data. J Stat Softw 62(7)
Jiménez-Gamero MD, Alba-Fernández M, Jodrá P, Barranco-Chamorro I (2017) Fast tests for the two-sample problem based on the empirical characteristic function. Math Comput Simul 137:390–410
Kim AY, Marzban C, Percival DB, Stuetzle W (2009) Using labeled data to evaluate change detectors in a multivariate streaming environment. Signal Process 89(12):2529–2536
Kirch C (2008) Bootstrapping sequential change-point tests. Seq Anal 27(3):330–349
Kuncheva LI (2013) Change detection in streaming multivariate data using likelihood detectors. IEEE Trans Knowl Data Eng 25(5):1175–1180
Lahiri SN (1999) Theoretical comparisons of block bootstrap methods. An Stat 27(1):386–404
Lahiri S (2003) Resampling methods for dependent data. Springer, New York
Lung-Yut-Fong A, Lévy-Leduc C, Cappé O (2015) Homogeneity and change-point detection tests for multivariate data using rank statistics. J Soc Franc Stat 156:133–162
Lütkepohl H (2005) New introduction to multiple time series analysis. Springer, New York
Ma TF, Yau CY (2016) A pairwise likelihood-based approach for changepoint detection in multivariate time series models. Biometrika 103(2):409–421
Matteson DS, James NA (2014) A nonparametric approach for multiple change point analysis of multivariate data. J Am Stat Assoc 109(505):334–345
Merlevède F, Peligrad M (2020) Functional clt for nonstationary strongly mixing processes. Stat Probab Lett 156:108581
Meynaoui A, Albert M, Laurent-Bonneau B, Marrel A (2019) Adaptive test of independence based on HSIC measures. arXiv:1902.06441
Nolan JP (2013) Multivariate elliptically contoured stable distributions: theory and estimation. Comput Stat 28(5):2067–2089
Politis DN, White H (2004) Automatic block-length selection for the dependent bootstrap. Econ Rev 23(1):53–70
Preuss P, Puchstein R, Dette H (2015) Detection of multiple structural breaks in multivariate time series. J Am Stat Assoc 110(510):654–668
Quessy JF, Éthier F (2012) Cramér-von Mises and characteristic function tests for the two and k-sample problems with dependent data. Comput Stat Data Anal 56(6):2097–2111
Quessy JF, Éthier F (2014) Two bootstrap strategies for a k-problem up to location-scale with dependent samples. J Probab Stat 2014:1–12
Rio E (2017) Asymptotic theory of weakly dependent random processes. Springer, New York
Rizzo ML, Székely GJ et al (2010) Disco analysis: a nonparametric extension of analysis of variance. Ann Appl Stat 4(2):1034–1055
Shao X (2010) The dependent wild bootstrap. J Am Stat Assoc 105(489):218–235
Spokoiny V et al (2009) Multiscale local change point detection with applications to value-at-risk. Ann Stat 37(3):1405–1436
Steland A, Von Sachs R et al (2017) Large-sample approximations for variance-covariance matrices of high-dimensional time series. Bernoulli 23(4A):2299–2329
Székely GJ, Rizzo ML (2005) Hierarchical clustering via joint between-within distances: extending Ward’s minimum variance method. J Classif 22(2):151–183
Székely GJ, Rizzo ML (2013) Energy statistics: a class of statistics based on distances. J Stat Plan Inference 143(8):1249–1272
Székely GJ, Rizzo ML (2016) Energy distance. Wiley Interdiscipl Rev 8(1):27–38
Tenreiro C (2019) On the automatic selection of the tuning parameter appearing in certain families of goodness-of-fit tests. J Stat Comput Simul 89(10):1780–1797
Yokoyama R (1980) Moment bounds for stationary mixing sequences. Z Wahrscheinlichkeitstheorie Verwandte Gebiete 52(1):45–57
Acknowledgements
The research was supported by Czech Science Foundation, Grant GAČR 18-08888S.
Author information
Authors and Affiliations
Corresponding author
Additional information
To the memory of Theophilos Cacoullos (1932–2020).
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Simos G. Meintanis—On sabbatical leave from the University of Athens.
Electronic supplementary material
Below is the link to the electronic supplementary material.
Appendix
Appendix
This section contains the proofs of the assertions formulated in Sect. 3. We start with a useful auxiliary relationship and auxiliary lemmas. D denotes a generic constant which may vary from line to line.
Direct calculations give for \(t=1,\ldots ,n\)
where
Next lemma summarizes needed assertions on partial sums of \(\alpha \)-mixing random variables.
Lemma 1
Let sequence \(\{({\varvec{x}}_t,{\varvec{y}}_t),\, t=1,\ldots \}\) be 2p-dimensional strictly stationary \(\alpha \)-mixing with coefficients \(\alpha (j)\) and let \(E h_t({\varvec{u}})=0\) then for arbitrary \({\varvec{u}}\in {\mathbb {R}}^p\) and any \(\xi >0\)
and for any \(b_1\ge \cdots \ge b_T>0\)
The inequalities remain true if \( h_t({\varvec{u}})\)’s is replaced by \( h_t({\varvec{u}}_1)- h_t({\varvec{u}}_2)\). If additionally
then for \(T\rightarrow \infty \) the limit distribution of \(\frac{1}{\sqrt{T}} \sum _{t=1}^T h_t({\varvec{u}})\) is asymptotically normal with zero mean and variance \(\sigma ^2\).
(b) Let assumption (A.1) and (25) be satisfied then
and
Proof
It follows from classical results on \(\alpha \)-mixing sequences of random variables, see e.g. Yokoyama (1980) and Doukhan (1994). \(\square \)
The following Lemma is substantial for the proof of Theorem 1.
Lemma 2
Let assumptions (A.1)–(A.2) and (25) be satisfied and let the null hypothesis hold true. Then
(a) for any compact subset F of \({\mathbb {R}}^p\) it holds
(b) there exist \(a>0\) and \(0<C<\infty \) such that for any \(0\le s\le 1\) and any \({\varvec{u}}_1\) and \({\varvec{u}}_2\) it holds
(c) the marginal distributions of \(\{Z_T(s, {\varvec{u}})\}\) converge to the marginal distributions of the Gaussian process \(\{Z(s, {\varvec{u}})\}\) with zero mean and covariance structure \((0\le s_1<s_2\le 1)\)
Proof
It follows the line of Lemma 7.1 in Hlávka et al. (2017), where the basic difference is the present paper works with functionals of partial sums of \(\alpha \)-mixing random vectors while in Hlávka et al. (2017) functionals of martingale differences are considered. In other words, instead of assertions on partial sums of martingale differences analogous results on partial sums of \(\alpha \)-mixing random vectors are applied.
By the assumptions for each \({\varvec{u}}\) the sequence \(\{ h_\tau ({\varvec{u}})\}_{\tau }\) is \(\alpha \)-mixing with the same coefficient as the sequence \(\{({\varvec{x}}_\tau , \varvec{y}_{\tau }\}_{\tau }\). Moreover,
By the assumptions \(E Z^2(s, {\varvec{u}})\) is uniformly bounded and therefore (a) holds true.
Concerning assertion (b) notice that
for some \(a>0\). We used the moment inequalities for \(\alpha \)-mixing sequences. Particularly, we used inequalities for \(\alpha \)-mixing.
Concerning (c) it is a consequence of Lemma 1. \(\square \)
Proof
(Theorem 1) From Lemma 7.1 in Hlávka et al. (2017) in addition to Theorem 22 in Ibragimov and Chasminskij (1981, pp. 380, 381) and Lemma 2 we get that for each \(s\in (0,1]\) as \(T\rightarrow \infty \)
for any compact \(F\subset {\mathbb {R}}^p\). Since \(W(\cdot )\) is integrable and \( EZ^2_T(s,{\varvec{u}})\le D\) then
which proves (a).
Next we have to study limit behavior of the process
Convergence finite dimensional distributions follow from Lemma 1 in combination with continuous mapping theorem. To obtain tightness by the Minkowski inequality we have
and further by the Jensen inequality for any \(\varepsilon >0\)
Then by Billingsley (2013, Theorem 15.6) we conclude that
which further implies that
\(\square \)
Proof
(Theorem 2) It follows the line of Theorem 1 and therefore is omitted. \(\square \)
Proof
(Theorem 3) By inequalities in Lemma 2 we notice that for any \(s\in (s_0,1]\)
and by assumptions
The proof can be straightforwardly finished. \(\square \)
Proof
(Theorem 4): It is omitted since it is done in the same way as Theorem 3. \(\square \)
Rights and permissions
About this article
Cite this article
Hlávka, Z., Hušková, M. & Meintanis, S.G. Change-point methods for multivariate time-series: paired vectorial observations. Stat Papers 61, 1351–1383 (2020). https://doi.org/10.1007/s00362-020-01175-3
Received:
Revised:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00362-020-01175-3