Pattern Recognition and Neural Networks

By B.D. Ripley

Cambridge University Press. ISBN 0 521 46086 7. January 1996.

[Image of Cover] Larger size

Hardback 246 x 189mm, 416 pages

Material available from here:
Description Contents Preface Datasets
Complements Errata Search References Download References
Reviews Contact author Publisher's Web Sites  

Pattern recognition has long been studied in relation to many different (and mainly unrelated) applications, such as

Human expertise in these and many similar problems is being supplemented by computer-based procedures, especially neural networks. Pattern recognition is extremely widely used, often under the names of `classification', `diagnosis' or `learning from examples'. The methods are often very successful, and this book explains why.

It is an in-depth study of methods for pattern recognition drawn from engineering, statistics, machine learning and neural networks. All the modern branches of the subject are covered, together with case studies of applications. The relevant parts of statistical decision theory and computational learning theory are included, as well as methods such as feed-forward neural networks (multi-layer perceptrons), radial basis functions, learning vector quantization and Kohonen's self-organizing maps. The theory explains how to tune and assess the methods.

This is a self-contained account, ideal both as an introduction for non-specialists and also as a handbook for the more expert reader.


Contents:

  1. Introduction and examples
  2. Statistical decision theory
  3. Linear discriminant analysis
  4. Flexible discriminants
  5. Feed-forward neural networks
  6. Non-parametric methods
  7. Tree-structured classifiers
  8. Belief networks
  9. Unsupervised methods
  10. Finding good features

The Preface is available on-line.

Some reviews are reproduced here.

The Kaufman & Rousseeuw software (pp. 311-312) is available as a single file of Fortran programs.

There is an Errata list available, and the references can be searched or retrieved.


Complements:

The `complements' provide an on-line updating of the book, both adding new references and ideas, and amplifying points in the light of readers' experiences. This is available as to be read on the Web or for downloading as PDF (for Acrobat Reader) or postscript or gzip-ed postscript.

This will be updated from time to time: the current version is 23 August 1996.


Author:

The author's email address is ripley@stats.ox.ac.uk and mail address is

Professor B.D. Ripley
Department of Statistics
1 South Parks Road
Oxford OX1 3TG
UK


Publisher:

Links are provided to CUP's home page and page for this book. In the USA they have home page and page for this book. You can order from those pages.


Last edited on Weds 11 Jun 1997 by Brian Ripley ripley@stats.ox.ac.uk