Information dynamics: foundations and applications (Q4520852)

From MaRDI portal

!

This is the item page for this Wikibase entity, intended for internal use and editing purposes. Please use the normal view instead:

scientific article; zbMATH DE number 1544062
Language Label Description Also known as
default for all languages
No label defined
    English
    Information dynamics: foundations and applications
    scientific article; zbMATH DE number 1544062

      Statements

      0 references
      0 references
      14 December 2000
      0 references
      univariate- and multivariate time series
      0 references
      information processing
      0 references
      parametric extraction of statistical structure
      0 references
      nonlinear chaotic dynamics
      0 references
      nonparametric characterization of time series
      0 references
      coding
      0 references
      spatiotemporal dynamical systems
      0 references
      spiking networks
      0 references
      Bayesian estimation
      0 references
      maximum-likelihood
      0 references
      maximum-entropy principle
      0 references
      minimum Kullback-Leibler entropy
      0 references
      feedforward neural networks
      0 references
      recurrent neural network
      0 references
      time series analysis
      0 references
      information theory
      0 references
      Information dynamics: foundations and applications (English)
      0 references
      The book deals with information dynamics, that is to say, information processing in complex dynamical systems. More explicitly, the authors ask ``under which circumstances it is possible to extract quantitatively and characterize the statistical structures underlying time-dependent real-world data'', and they propose a solution to these problems via a unifying approach (based on information theory) to the extraction of parametric, nonparametric and semiparametric statistical structure. What is striking in this book is that it does not contain any theorem with proof. This is due to the fact that it mainly summarizes some results obtained in an industrial project which obviously had to work in the applications irrespective of any theory. It comprises 10 chapters including the introduction. In order to prepare the reader, Chapter 2 (dynamical systems: an overview) provides an overview of dynamical systems, including chaotic deterministic systems, stochastic systems and time series. The problem of parametric extraction of statistical structure is described in Chapter 3 (statistical structure extraction in dynamical systems: parametric formulation), and a solution is proposed which is based on a combination of information theory with neural networks. Chapter 4 (applications: parametric characterization of time series) applies the approach to the modelling of time series related to nonlinear chaotic dynamics. Chapter 5 (statistical structure extraction in dynamical systems: nonparametric formulation) deals with an information theoretic nonparametric formulation for statistical structure extraction in dynamical systems. The concept of information flow is introduced. Chapter 6 (applications: nonparametric characterization of time series) applies the method described in Chapter 5 to the nonparametric characterization of time series. In Chapter 7 (statistical structure extraction in dynamical systems: semiparametric formulation), the purpose of the authors is to detect the Markovian features of dynamical systems by analyzing their flow of information. Univariate-and multivariate time series are considered. In Chapter 8 (applications: semiparametric characterization of time series), the method of the preceeding chapter is applied to the semiparametric characterization of univariate and multivariate time series. Chapter 9 (information processing and coding in spatiotemporal dynamical systems: spiking networks) considers distributed parameter systems, and a principle for the generation of spiking coding and information processing is stated. Chapter 10 (applications: information processing and coding in spatiotemporal dynamical systems) applies results of Chapter 9 to neural networks. The basic statistical tools used are summarized in Chapter 3: Bayesian estimation, maximum-likelihood, maximum-entropy principle, minimum Kullback-Leibler entropy, feedforward neural networks, recurrent neural network. In the opinion of the reviewer, the (or one of the) main interests of the book is to provide a new approach to time series analysis via information theory only.
      0 references

      Identifiers

      0 references
      0 references
      0 references
      0 references
      0 references
      0 references
      0 references