2011, Vol.14, No.1, pp.102-105
In the paper "Information Geometry of a regime-switching model
with time-varying parameters" [1], we discussed the
possibility to use tools derived from Information Geometry in
order to analyze dynamical systems in which the motion randomly
switches between two regimes using Markovian transitions. We
proved that, in order to understand when an change can appear, the
network properties of the system are important because they
accelerate the diffusion of the information. In this note we
recall the most important results obtained announcing some
possible applications to detect the regime-switching in real
phenomena.
Key words:
regime-switching models, information geometry,
neural networks, (EM) and (em) algorithms
Full text: Acrobat PDF (196KB)
Copyright © Nonlinear Phenomena in Complex Systems. Last updated: May 5, 2011