Preprint Article Version 1 This version is not peer-reviewed

Representation and Characterization of Nonstationary Processes by Dilation Operators and Induced Shape Space Manifolds

Version 1 : Received: 21 July 2018 / Approved: 23 July 2018 / Online: 23 July 2018 (06:02:32 CEST)

A peer-reviewed article of this Preprint also exists.

Dugast, M.; Bouleux, G.; Marcon, E. Representation and Characterization of Nonstationary Processes by Dilation Operators and Induced Shape Space Manifolds. Entropy 2018, 20, 717. Dugast, M.; Bouleux, G.; Marcon, E. Representation and Characterization of Nonstationary Processes by Dilation Operators and Induced Shape Space Manifolds. Entropy 2018, 20, 717.

Journal reference: Entropy 2018, 20, 717
DOI: 10.3390/e20090717

Abstract

We proposed in this work the introduction of a new vision of stochastic processes through geometry induced by dilation. The dilation matrices of a given process are obtained by a composition of rotation matrices built in with respect to partial correlation coefficients. Particularly interesting is the fact that the obtention of dilation matrices is regardless of the stationarity of the underlying process. When the process is stationary, only one dilation matrix is obtained and it corresponds therefore to Naimark dilation. When the process is nonstationary, a set of dilation matrices is obtained. They correspond to Kolmogorov decomposition. In this work, the nonstationary class of periodically correlated processes was of interest. The underlying periodicity of correlation coefficients is then transmitted to the set of dilation matrices. Because this set lives on the Lie group of rotation matrices, we can see them as points of a closed curve on the Lie group. Geometrical aspects can then be investigated through the shape of the obtained curves, and to give a complete insight into the space of curves, a metric and the derived geodesic equations are provided. The general results are adapted to the more specific case where the base manifold is the Lie group of rotation matrices, and because the metric in the space of curve naturally extends to the space of shapes, this enables a comparison between curves’ shapes and allows then the classification of processes’ measure.

Subject Areas

nonstationary processes; spectral measure; differential geometry; shape manifold; square root velocity function; Lie group

Comments (0)

We encourage comments and feedback from a broad range of readers. See criteria for comments and our diversity statement.

Leave a public comment
Send a private comment to the author(s)
Views 0
Downloads 0
Comments 0
Metrics 0


×
Alerts
Notify me about updates to this article or when a peer-reviewed version is published.