on June 12, 2004
Without any hesitation, I consider this book as a masterpiece in the area of statistical signal processing. Kay takes the reader to the journey of estimation theory as if a science teacher takes his students to a field trip. The one special feature of this book is the convergence of thought that reader obtains upon reading the book. Kay lays a fundamental bridge between various estimators using his succinct style for describing the subject.
Few special areas require more attention in this book. For example the coverage of EM methods is very condense and requires more elaboration. Also there is no discussion on the estimation methods using higher order statistics.
Overall I consider this book as the best book I have read ever and I highly recommend this book to those who want to obtain an ever-lasting view on statistical signal processing.
on August 11, 2003
I've had tough courses on statistical signal processing as a post-grade student. I am often confused in front of a problem and turning back to the notes taken in class doesn't help much.
When you read this book all gets bright. I am still wondering how some teachers can be so confusing while such good books do exist...
However don't count on it for in depth mathematical demonstrations, it starts with a practical problem and explains how to model things. Thus it is a bit bottom-up but anyway starting from a good graduate level in signal and stats.
I got this one at the library but already ordered a copy for myself and am planning to get part2 on detection.
on August 9, 2000
In this book, Steven M. Kay has produced an excellent tutorial and research reference book on estimation theory. The book covers enough introductory material for someone with a reasonable undergraduate understanding of statistics to pick up the ideas quickly. The theory is illustrated with very concrete examples; the examples give an "under-the-hood" insight into the solution of some common estimation problems in signal processing. If you're a statistician, you might not like this book. If you're an engineer, you will like it.
on June 8, 2003
This text is very good for those who start doing research in statistical signal processing. A lot of explanations, technical terms are well presented and consistent, plus a number of examples that help you to learn about different statistical signal processing concepts and algorithms. Research students can be beneficial alot from this text.