patternrecognition.co.za

Home | News | Services | Downloads | Forum | About us | Contact us


 


 Resources
 
 Publications
 Conference papers

 Source Code
 Classifier source code

 Tutorials 
 Pattern recognition tutorials

 
Terminology
 Definitions of pattern
 recognition terms

 
Pattern Recognition
 Applications 
 A summary of pattern
 recognition applications

 Artificial Intelligence 
 Interesting articles and  
 discussions regarding
 artificial intelligence


 Classification Applet
 
 Applet
 Online implementation of
 various classifiers

 
Data Set Format
 Description of the data set
 format used for the
 classification applet


 
Example Data Sets
 Downloadable data set
 examples

 
Classification Applet
 Documentation
 Description of the algorithms
 used


The purpose of this web site is to provide resources to  experienced researchers as well as new comers in the fields of pattern recognition, artificial intelligence, machine learning and other overlapping research fields.

The focus of this site's research is to
fully describe the relationships between data characteristics and classifier behaviour, and to develop algorithms that automatically select classifiers and parameters appropriate for a given set or subset of data. Relevant publications are available from the 'Publications' link and other relevant tutorials are available at the 'Tutorials' link.

It is clear from the literature that there is no best classifier for all types of problems. Some guidelines have been proposed in the literature for classifier selection. These guidelines, however, do not provide much insight on the specific characteristics of the data that will determine the preference of classifier.
An empirical comparison of classifiers on any problem can be performed by using the classification applet available at the 'Applet' link.

Pattern recognition is a very active research area which overlaps with various other research fields such as Machine Learning, Artificial Intelligence, Data Mining, Probability Theory, Algebra and Calculus. Source code for various classifiers and other numerical and statistical algorithms are available at the 'Source Code' link. Terminology used in these research fields can be found at the 'Tutorials' link.

Classifiers cover a wide range of information processing problems. These problems often have great importance and include speech recognition, classification of handwritten characters, fault detection in machinery, medical diagnosis and many other. A non-exhaustive list of pattern recognition applications is available at the 'Pattern Recognition Applications' link'.

Classification algorithms have been applied to various comercial products very succesfully. Commercial Neural Network Classification, Regression, Data Analysis and many other applications are available at the links on the left hand side.


































Recommended Links

UCI Machine Learning Repository 

Repository of various classification data sets

Meta-Level Learning
Web site dedicated to meta level learning

Weka
Machine learning package

Yale
Machine learning package 

Statistical Pattern Recognition Toolbox
Matlab pattern recognition toolbox

libSVM
SVM classification package

Pattern Recognition Tools
Matlab pattern recognition toolbox

Opt Neural Network
Neural network source code


Copyright 2007. www.patternrecognition.co.za