Abstract
Non-stationarity time series are very common in physical, biological and in real-world systems in general, ranging from geophysics, econometrics or electroencephalography to logistics. Identifying, detecting and adapting learning algorithms to non-stationary environments is a fundamental task in many data mining scenarios; however it is often a major challenge for current methodologies. Data analysis in the context of time-varying statistical moments is a very active research direction in machine learning and in computational statistics; but theoretical insights into latent causes of non-stationarity in empirical data are very scarce. In this study, we evaluate the capacity of the trajectory classification error statistic in order to detect a significant variation in the underlying dynamics of data collected in multiple stages. We analysed qualitatively the conditions leading to observable changes in non-stationary data generated by Duffing non-linear oscillators; which are ubiquitous models of complex classification problems. Analyses are further benchmarked in a dataset consisting of atmospheric pollutants time series.
This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsPreview
Unable to display preview. Download preview PDF.
References
Balaguer-Ballester, E., Tabas-Diaz, A., Budka, M.: Can we identify non-stationary dynamics of trial-to-trial variability? PLoS ONE 9, e95648 (2014)
Balaguer-Ballester, E., Camps-Valls, G., Carrasco-Rodriguez, J.L., Soria-Olivas, E., del Valle-Tascon, S.: Effective one-day ahead prediction of hourly surface ozone concentrations in Eastern Spain using linear models and neural networks. Ecological Modelling 156, 27–41 (2002)
Balaguer-Ballester, E., Lapish, C., Seamans, J., Durstewitz, D.: Attracting dynamics of frontal cortex ensembles during memory guided decision making. PLoS Computational Biology 7, e1002057 (2011)
Beck, J.M., Ma, W.J., Pitkow, X., Latham, P.E., Pouget, A.: Not noisy, just wrong: The role of suboptimal inference in behavioral variability. Neuron 74, 33–39 (2012)
Blythe, D.A.J., von Bunau, P., Meinecke, F.C., Robert-Muller, K.: Feature extraction for change-point detection using stationary subspace analysis. IEEE Trans. Neural Networks 23, 631–643 (2012)
Bouchachia, A.: Incremental learning with multi-level adaptation. Neurocomputing 74, 1785–1799 (2011)
Chen, J.S., Hu, N.J.: A frequency domain test for detecting nonstationary time series. Computational Statistics and Data Analysis (in press, 2014)
Du, J., Cui, M.: Solving the forced duffing equation with integral boundary conditions in the reproducing kernel space. International Journal of Computer Mathematics 87, 2088–2100 (2010)
Durstewitz, D., Balaguer-Ballester, E.: Statistical approaches for reconstructing neuro-cognitive dynamics from high-dimensional neural recordings. Neuroforum 1, 89–98 (2010)
Feng, Z., Chen, G., Hsu, S.: A qualitative study of the damped duffing equation and applications. American Institute of Mathematical Sciences 6, 1097–1112 (2006)
Gomez-Sanchis, J., Martin-Guerrero, J.O., Soria-Olivas, E., Vila-Frances, J., Carrasco, J.L., del Valle-Tascon, S.: Neural networks for analysing the relevance of input variables in the prediction of tropospheric ozone concentration. Atmospheric Environment 40, 6173–6180 (2006)
Haraa, S., Kawaharaa, Y., Washioa, T., von Bnau, P., Tokunagac, T., Yumotod, K.: Separation of stationary and non-stationary sources with a generalized eigenvalue problem. Neural Networks 33, 7–20 (2012)
Kuncheva, L.I., Rodriguez, J.J.: Interval feature extraction for classification of event-related potentials (erp) in eeg data analysis. Progress in Artificial Intelligence 2, 65–72 (2012)
Masquelier, T.: Neural variability, or lack thereof. Frontiers in Comput. Neurosci. 7, 1–7 (2013)
Nason, G.: A test for second-order stationarity and approximate confidence intervals for localized autocovariances for locally stationary time series. J. R. Statist. Soc. B 75, 879–904 (2013)
Priestley, M.B., Subba, T.: A test for non-stationarity of time-series. J. R. Statist. Soc. B 31, 140–149 (1969)
Sauer, T., Yorke, J., Casdagli, M.: Embedology. J. Stat. Phys. 65, 579–616 (1992)
Sayed-Mouchaweh, M., Lughofer, E. (eds.): Learning in Non-Stationary Environments. Springer (2012)
Scholkopf, B., Smola, A.J.: Learning with kernels. MIT Press (2002)
von Bunau, P., Meinecke, F.C., Kiraly, F.J., Robert-Muller, K.: Finding stationary subspaces in multivariate time series. Phys. Rev. Lett. 103, 214101 (2009)
Wiggins, S.: Introduction to applied nonlinear dynamical systems and chaos. Springer (2013)
Zliobaite, I., Bifet, A., Pfahringer, B., Holmes, G.: Active learning with drifting streaming data. IEEE Transactions on Neural Networks and Learning Systems (2013) (in press)
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2014 Springer International Publishing Switzerland
About this paper
Cite this paper
Balaguer-Ballester, E., Tabas-Diaz, A., Budka, M. (2014). Empirical Identification of Non-stationary Dynamics in Time Series of Recordings. In: Bouchachia, A. (eds) Adaptive and Intelligent Systems. ICAIS 2014. Lecture Notes in Computer Science(), vol 8779. Springer, Cham. https://doi.org/10.1007/978-3-319-11298-5_15
Download citation
DOI: https://doi.org/10.1007/978-3-319-11298-5_15
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-11297-8
Online ISBN: 978-3-319-11298-5
eBook Packages: Computer ScienceComputer Science (R0)