Over multiple rule-blocks to modular nets
Lambert Spaanenburg, Walter Jansen, Jos A. G. Nijhuis · 2002
Real production data are vague and irreproducible. This suggests fuzzy knowledge acquisition for an on-line learned, combined fuzzy/neural network. This paper advocates a modular neural-only network based on the injection of knowledge from a multiple rule-block fuzzy specification. A typical controller rakes 10 - 25 blocks with 10 rules on average, which after injection in a neural net can be personalized and adapted by small (typical 100) measurement sets.