Classification, Parameter Estimation and State Estimation: An Engineering Approach Using MATLAB

Appendix E

Introduction to PRTools

E.1 MOTIVATION

In statistical pattern recognition we study techniques for the generalization of decision rules to be used for the recognition of patterns in experimental data sets. This area of research has a strong computational character, demanding a flexible use of numerical programs for data analysis as well as for the evaluation of the procedures. As new methods keep being proposed in the literature, a programming platform is needed that enables a fast and flexible implementation of such algorithms. Because of its widespread availability, its simple syntax and general nature, MATLAB is a good choice for such a platform.

The pattern recognition routines and support functions offered by PRTools represent a basic set covering largely the area of statistical pattern recognition. Many methods and proposals, however, are not yet implemented. Neural networks are only implemented partially, as MATLAB already includes a very good toolbox in that area. PRTools has a few limitations. Due to the heavy memory demands of MATLAB, very large problems with learning sets of tens of thousands of objects cannot be handled on moderate machines. Moreover, some algorithms are slow as it can be difficult to avoid nested loops. In the present version, the handling of missing data has been prepared, but no routines are implemented yet. The use of fuzzy or symbolic data is not supported, except for soft (and thereby also fuzzy) labels which are used by just a few routines. ...