Item Successfully Added to Cart
An error was encountered while trying to add the item to the cart. Please try again.
Please make all selections above before adding to cart
The following link can be shared to navigate to this page. You can select the link to copy or click the 'Copy To Clipboard' button below.
Copy To Clipboard
Successfully Copied!
Codage Universel et Identification d’ordre par Sélection de Modèles

Elisabeth Gassiat University Paris-Sud, Orsay, France
A publication of the Société Mathématique de France
Available Formats:
Softcover ISBN: 978-2-85629-782-7
Product Code: COSP/21
List Price: $67.00 AMS Member Price:$53.60
Please note AMS points can not be used for this product
Click above image for expanded view
Codage Universel et Identification d’ordre par Sélection de Modèles
Elisabeth Gassiat University Paris-Sud, Orsay, France
A publication of the Société Mathématique de France
Available Formats:
 Softcover ISBN: 978-2-85629-782-7 Product Code: COSP/21
 List Price: $67.00 AMS Member Price:$53.60
Please note AMS points can not be used for this product
• Book Details

Cours Spécialisés
Volume: 212014; 140 pp
MSC: Primary 62; 68;

A note to readers: This book is in French.

These notes are at the interface between information theory and statistics. They highlight how universal coding and adaptive compression are linked with the statistical inference of random processes, by maximum likelihood or Bayesian methods.

The author starts with classic tools for dealing with finite alphabets, then presents the recent theory of universal coding in infinite aphabets. She shows how it is used to solve problems of order identification, in particular for hidden Markov models.

Postgraduate students and research mathematicians interested in universal coding and order identification.

• Request Review Copy
Volume: 212014; 140 pp
MSC: Primary 62; 68;

A note to readers: This book is in French.

These notes are at the interface between information theory and statistics. They highlight how universal coding and adaptive compression are linked with the statistical inference of random processes, by maximum likelihood or Bayesian methods.

The author starts with classic tools for dealing with finite alphabets, then presents the recent theory of universal coding in infinite aphabets. She shows how it is used to solve problems of order identification, in particular for hidden Markov models.