SOTA -- A visualisation tool for ART neural networks

CS-TR-93-3

Authors: Guszti Bartfai, Martin Elliffe, Peter Wood
Pages: 7
Source: GZipped PostScript (92kb); Adobe PDF (284kb)


This paper introduces the prototype of a visualisation tool (SOTA) for Adaptive Resonance Theory (ART) neural networks. The tool is designed to provide novel views that help researchers understand the dynamic processes inherently present in ART networks, as a result of the interaction between the two layers and subsystems of ART. Two such views, circle diagrams and input-relative diagrams, are discussed in detail. Examples are also given showing some features of these views and SOTA itself.

[Up to Computer Science Technical Report Archive: Home Page]