Time Series Analysis by State Space Methods: Second Edition (Oxford Statistical Science Series)

This is the last book by "Mr. Time Series." Durbin knew everyone involved in the development of modern statistical analysis of time series. This effort, written with Koopmans (of Commandeur and Koopmans) is a graduate-level presentation of state space methods, whereas the Commandeur/Koopmans effort can be shared with good undergraduates.

Is a excellent articule

I think it's one of best books in state space model. It comes in 2012 and covers a lot of updates in the field. A few problems,1. not too many examples, so if you are new to SSM, it may be difficult for you to understand, and you need other books accompany, likeDLM with R, etc.2. This books' perspective is a bit from other books. In Dynamic Linear Model with R and the other text by Mike West, the treatment are similar3. Detailed algorithm is not included. And the code is not free to use.But overall, this is a very good book who anyone who want to be serious with State Space Models.

If you're on the hunt for a comprehensive and detailed mathematical treatment of State Space modeling, this book may be what you're looking for. It's a "heavy" textbook, not a "how-to" cookbook, but is well-organized and well-written. The first author was James Durbin, the renowned statistician who passed away in 2012 at the age of 88. His frequent collaborator, Siem Jan Koopman, is widely published on time series analysis and econometrics topics.

Good book to learn how to do filtering. clear and concise. I havent finished the book but I believe they could add more on high dimensional problems.

Excellent treatment for the frontier research in time series with emphasis on MCMC with state-space approach.

This is by far the most comprehensive work on linear state space modeling. I frequently use it as a reference. It's exposition of the topic is clear and helped me learn the material from a point of relative ignorance.

"Preface to Second Edition", found via "Search inside", discusses the changes since the first edition, and these do address readers' comments on that book's Amazon page, by expanding coverage of the non-linear/non-normal case. I must say the book still feels like one about Kalman filter, but page-count comparisons* I now invoke to justify this feeling may be misleading. "Time series analysis by state space methods" was not quite what I was looking for - I'd prefer something less dry/technical, and more application-minded and hands-on (regrettably, readers' complaints regarding companion software have not been heeded; this, and the book's steep price, are my excuse not to give it five stars) - and I did not dig deep, but the overall impression is that of a comprehensive, rigorous and reasonably compact exploration of the field.* The "non-linear/non-normal" Part II has half as many pages as the "linear/normal" Part I; linearity gains extra ground in Part II's chapter on approximate methods (including extended and unscented Kalman filer), and leaves less space for particle filtering. The latter gets a total of 23 pages; I opted to read the survey paper by Drew Creal, available online, and to revisit the concluding chapter of "Dynamic linear models with R" by Petris et al.

Il libro tratta l'argomento in maniera esaustiva.Come sempre con Amazon le condizioni del libro non sono mai ottimali.

Delivered as claimed.

Libro di livello avanzato sui modelli state space. Da un certo punto le spiegazioni si complicano e ci sono rimandi a capitoli successivi. Mi sembra comunque migliorato rispetto all'edizione precedente.