About this book
Since 1975, "The Analysis of Time Series: An Introduction" has introduced legions of statistics students and researchers to the theory and practice of time series analysis. With each successive edition, best-selling author Chris Chatfield has honed and refined his presentation, updated the material to reflect advances in the field, and presented interesting new data sets. The sixth edition is no exception. It provides an accessible, comprehensive introduction to the theory and practice of time series analysis. The treatment covers a wide range of topics, including ARIMA probability models, forecasting methods, spectral analysis, linear systems, state-space models, and the Kalman filter. It also addresses nonlinear, multivariate, and long-memory models.
The author has carefully updated each chapter, added new discussions, incorporated new datasets, and made those datasets available for download.This sixth edition includes: a new section on Handling Real Data; new discussion on prediction intervals; a completely revised and restructured chapter on more advanced topics, with new material on the aggregation of time series, analyzing time series in finance, and discrete-valued time series; a new chapter of Examples and Practical Advice; and, thorough updates and revisions throughout the text that reflect recent developments and dramatic changes in computing practices over the last few years. The analysis of time series can be a difficult topic, but as this book has demonstrated for two-and-a-half decades, it does not have to be daunting.
Simple descriptive techniques - types of variation, stationary time series, the time plot, transformations, analyzing series which contain a trend, which contain seasonal variation, autocorrelation, other tests of randomness; probability models for time series - stochastic and stationary processes, the autocorrelation function; the Wold decomposition theorem; estimation in the time domain - fitting an autoregressive process and a moving average process, estimating the parameters of an ARMA model and an ARIMA model, the Box-Jenkins seasonal model, residual analysis; forecasting - univariate and multivariate procedures, a comparative review of forecasting procedures, prediction theory; stationary processes in the frequency domain - the spectral distribution function and density function; the spectrum of a continuous process, derivation of selected spectra; spectral analysis - Fourier analysis, a simple sinusoidal model;m periodogram analysis, a comparison of different estimation procedures; bivariate process - cross-covariance and cross-correlation functions, the cross-spectrum; linear systems - linear systems in the time domain and in the frequency domain, identification of linear systems; state-space models and the Kalman filter; some other topics - control theory, modelling non-stationary series, non-linear models, model identification tools, autoregressive spectrum estimation, spatial series, crossing problems, observations at unequal intervals. Appendices: the Fourier, Laplace and Z transforms; the Dirac delta function; covariance; some worked examples.