README

Hierarchical-Neural-Tree

Hierarchical classification and vector quantization with neural trees in pascal

This was my Diploma (similar to Master) thesis in psychology. Implemented around 2000 in Borland Delphi, that was based on the programming language Pascal. Today it is known as Delphi by Embarcadero. The projects aim was to make hierarchical neural tree networks accessible to cognitive psychologists. At that time they were toying around with classification problems and mainly used Kohonen nets. Which are neat but they did not support a hierachical stucture of classification. So this should have enabled psychologists with very little programming skills to use this class of networks.

Based on: Li, T., Fang, L., & Li, K. Q. (1993). Hierarchical classification and vector quantization with neural trees. Neurocomputing, 5(2), 119-139.

The Windows EXE 'Neuraler_Baum.exe' runs in Windows 10, however the provided trainig examples 'abc.db' and 'ordner1.db' require the Borland Paradox Database Engine to be processed.

Code on! Ferenc

Regensburg, Germany, 2016-04-16

PS: If anyone feels compelled to translate the Germany content into English, you will have my eternal gratitude!