
Author(s) 
Gani, J. (ed.) Priestley, M.B. (ed.) 

Title  Essays in time series and allied processes  Papers in honour of E.J. Hannan 
Publisher  Applied probability trust, University of Sheffield 
Year of publication  1986 
Reviewed by  Günter Vojta 
Time series are realizations of stochastic processes observed at discrete points of time. The need of time series analysis is present in almost every scientific field. The theory of time series constitutes nowadays a particularly important and widely developed area of stochastics with many applications in physics, chemistry, biology, in the technical sciences and in economy, among others. Therefore, books like that under review are welcome.
The volume is edited as a "festschrift" on the occasion of the 65th birthday of E.J. Hannan, Professor of Statistics at the Australian National University, Canberra, a wellknown leading authority on time series. It contains 32 papers of very different length arranged in 7 parts: structure and general methods for time series, estimation for time series, hypothesis testing and distribution theory for time series, nonlinear and nonstationary systems in time series, random fields and point processes, allied stochastic processes, and, finally, algorithms and computations.
The topics treated range from linear dynamic systems and their spectral analysis to mixed Cox processes and to random cyclic transformations of points. Applications are given e.g. to the analysis of earthquake data, to accident statistics and to population processes. Included is a list of (more than 100) publications by E.J. Hannan. An index of contributors to the book and a subject index (3 pp.) supplement the text.
According to the style of the Journal of applied probability, each of the papers is well accessible due to an abstract and keywords at the beginning. The book is altogether of a high standard. It represents a good source of professional information for all those engaged in time series and applied probability theory.