Time Series: Theory and Methods

by Peter J. Brockwell and Richard A. Davis

Book Overview

This edition contains a large number of additions and corrections scattered throughout the text, including the incorporation of a new chapter on state-space models. The companion diskette for the IBM... This description may be from another edition of this product.

• Format: Hardcover
• Language: English
• ISBN: 0387964061
• ISBN13: 9780387964065
• Release Date: January 1987
• Publisher: Springer
• Length: 519 Pages
• Weight: 2.00 lbs.

Customer Reviews

Great book

I reviewed this book once before under the pen name statman13. So look at that review to get most of my thoughts about it. I taught times series analysis as a graduate course at UC Santa Barbara many years ago. That was long before this book came out. I used Wayne Fuller's book as a text because it had balanced coverage of time domain and frequency domain approaches. If I were to do it over today I would use Brockwell and Davis' book as it has the right level of theory and also a proper mix of frequency and time domain. I know Richard Davis to be an excellent probabilist and very knowledgeable about stochastic process. I collaborated with him on a paper in extreme value theory. I also had the privilege of refereeing one of his early papers on extreme values that was part of his disseration and was eventually published in the Annals of Probability.

Time Series: Theory and Methods

Excellent reading. This book covers mainly the frequentist approach to time series analysis in a very informative way. The book starts off by introducing Hilbert spaces, then moves to stationary ARMA processes and so on. My favourite is chapter 10, Inference for the Spectrum of a Stationary Process, in which different tests are considered for periodicities at known and unknown frequencies.

Rigorous, difficult, but feasible

Of course, this an advanced textbook on Time Series. The reader is supposed to have been introduced to the subject, and certainly is looking for a more theoretical treatment.If you want to learn time series for the first time, this is not the book.If you want a friendly book, do not see springer's publications.However, if you want a fair rigourous book, you have found it.I think the exercises are illustrative, but sometimes long.