eduOptim ver 0.9

Links

Eduoptim 2

A new version of Eduoptim of which I am a co-author.

NEOS Guide Optimization Tree

Part of NEOS Guide introducing different areas of optimization and most important algorithms.

The program illustrates a few numerical optimization methods. Individual steps of finding the minimum with selected algorithm are shown both in numerical and graphical form.

Minimized functions can include theoretically any number of variables and can use most mathematical functions included in standard C library.

The following optimization methods are implemented:

steepest descent,
Fletcher-Reeves,
Polak-Ribiera¹,
Newton's¹.

Descent methods (first three) use one of two line-search algorithms:

α-division,
quadratic approximation¹.

Adding new methods shouldn't pose any problems (except methods like simplex that would require redesigning of the plotting module).

The interface is Polish only (but see the README and english.png files).

¹ these methods and a few other program functions have been written by Karol Matuszkiewicz and Mateusz Mucha. More information in the README file.

2006