Analyse de données chronologiques. (French) Zbl 0564.62069

Séminaire de Mathématiques Supérieures. Séminaire Scientifique OTAN (NATO Advanced Study Institute), 89. Département de Mathématiques et de Statistique, Université de Montréal. Montréal: Les Presses de l’Université de Montréal. 169 p. $ 20.00 (1985).
This book approaches the analysis of second-order stochastic processes, in particular of ARMA processes, through the geometric representation in a Hilbert space.
Chapter 1 sets out the geometric approach in the framework of time series analysis. In chapter 2 the basic definitions of second-order stochastic processes are presented. Chapters 3 and 4 deal with ARMA processes with time-dependent coefficients (called evolutionary ARMA processes), and ARMA processes with constant coefficients, respectively.
Chapter 5 discusses the Box and Jenkins approach to time series analysis. In chapter 6 the estimation of ARMA models is treated via the maximum likelihood principle and finally the generalized ARMA models are considered in chapter 7.
Reviewer: P.A.Morettin


62M10 Time series, auto-correlation, regression, etc. in statistics (GARCH)
62-02 Research exposition (monographs, survey articles) pertaining to statistics