Identification of generic fuzzy models of the human Central Nervous System control

The aim of this work is to identify generic (multipatient) models of the human Central Nervous System (CNS) control by means of the Fuzzy Inductive Reasoning (FIR) methodology. Previous works on this direction showed that FIR approach was able to capture in a reliable way the dynamic behavior of the CNS control for 5 different patients. For each one of them a fuzzy inductive model was identified. The errors obtained in the validation process were lower than 1%. From the medical perspective, it is very useful to have an available generic model of the CNS control that represents a specific class of patients. In this paper, two different approaches were used that allowed us to obtain a fully generic model and a generic structure model. These models are compared from both the prediction accuracy and the generalization power point of view.