Author Description

This third edition of Time Series is a thorough revision of the classic text by the late Sir Maurice Kendall. The subject has undergone dramatic changes in the last 20 years, and Keith Ord here presents an up-to-date treatment that uses some of the classical methods, such as data-analytic devices, before considering modern approaches to model building. The mathematics is kept at a reasonable level to make the book accessible to undergraduates and postgraduate researchers in many applied fields, such as economics, heliology, oil prospecting, and management science. Both ARIMA and structural approaches to model building are discussed, and the work now includes worked examples using real and simulated data. End-of-chapter exercises have been added, along with 16 data sets for further analysis. The text also offers a discussion of spectral methods, including fast Fourier transforms, intervention analysis, and transfer function models. An introduction to multiple time series is provided, and new sections describe traditional forecasting procedures, comparative evaluations, and automatic model selection procedures.