Rule Extraction from a Multilayer Feedforward Trained Network via Interval Arithmetic Inversion | Publicación