The use of methods of time series analysis has become of increased interest in recent years. Although the methods are rather well developed and understood for univariate time series analysis, the situation is not so complete for the multivariate case. This book is designed to introduce the basic concepts and methods that are useful in the analysis and modeling of uni- and multivariate time series, with illustration of these basic ideas. The development includes both traditional topics such as autocovariances and autocorrelation functions and matrices of stationary processes, properties of scalar and vector AR, MA and ARMA models, forecasting, least squares and maximum likelihood estimation techniques, and model checking diagnostics for residuals, as well as topics of more recent interest for vector ARMA models such as canonical correlation analysis, state-space models and Kalman filtering techniques and applications.
Five classes of model, which can be used to describe a spectrum of practical situations, ranging from the simple to the more complex: univariate models, transfer function models, intervention models, multivariate models and multivariate transfer function models, are discussed in the book.
The purposes of this book are (1) to present the concepts of uni- and multivariate time series modeling and forecasting in a manner that is friendly to the reader lacking a sophisticated background in mathematical statistics, and (2) to help the reader learn the art of time series modeling by means of detalied case studies. Chapters 1-8 present the essential concepts underlying the time series analysis and modeling, and Chapters 9-10 contain practical rules to guide the analyst, along with case studies showing how the techniques and methods are applied. Listings of the data sets used in the numerical examples are included in Appendix A, and a short description of the TS-System software package, developed by the author and used in the case studies, is given in Appendix B.
The book is intended for students, researchers and practitioners who have had little statistical training but either want to study these time series analysis methods or are confrunted with decisions based on these methods. Of course, those with more rigorous statistical training, but unfamiliar with time series analysis methods will certainly benefit from studying this book.
No matter who makes use of this book, it is important to remember that its orientation is an application one and the readers are strongly encouraged to develop a hands-on experience using real data encountered in their respective fields of study or work.