Show simple item record

dc.contributor.authorLacasa, L
dc.date.accessioned2017-01-04T14:05:11Z
dc.date.issued2014-07-29
dc.date.submitted2016-12-16T12:56:26.623Z
dc.identifier.urihttp://qmro.qmul.ac.uk/xmlui/handle/123456789/18346
dc.description.abstractDynamical processes can be transformed into graphs through a family of mappings called visibility algorithms, enabling the possibility of (i) making empirical data analysis and signal processing and (ii) characterising classes of dynamical systems and stochastic processes using the tools of graph theory. Recent works show that the degree distribution of these graphs encapsulates much information on the signals variability, and therefore constitutes a fundamental feature for statistical learning purposes. However, exact solutions for the degree distributions are only known in a few cases, such as for uncorrelated random processes. Here we analytically explore these distributions in a list of situations. We present a diagrammatic formalism which computes for all degrees their corresponding probability as a series expansion in a coupling constant which is the number of hidden variables. We offer a constructive solution for general Markovian stochastic processes and deterministic maps. As case tests we focus on Ornstein-Uhlenbeck processes, fully chaotic and quasiperiodic maps. Whereas only for certain degree probabilities can all diagrams be summed exactly, in the general case we show that the perturbation theory converges. In a second part, we make use of a variational technique to predict the complete degree distribution for special classes of Markovian dynamics with fast-decaying correlations. In every case we compare the theory with numerical experiments.
dc.language.isoenen_US
dc.subjectnlin.CD
dc.subjectnlin.CD
dc.subjectphysics.data-an
dc.titleOn the degree distribution of horizontal visibility graphs associated to Markov processes and dynamical systems: diagrammatic and variational approaches
dc.typeJournal Article
dc.rights.holder© 2014 IOP Publishing Ltd & London Mathematical Society
dc.identifier.doi10.1088/0951-7715/27/9/2063
pubs.author-urlhttp://arxiv.org/abs/1402.5368v1
pubs.organisational-group/Queen Mary University of London
pubs.organisational-group/Queen Mary University of London/Faculty of Science & Engineering
pubs.organisational-group/Queen Mary University of London/Faculty of Science & Engineering/Mathematical Sciences - Staff and Research Students
pubs.publisher-urlhttp://dx.doi.org/10.1088/0951-7715/27/9/2063


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record

Return to top