MIXED GRAPHICAL MODELS FOR SIMULTANEOUS MODEL IDENTIFICATION AND CONTROL APPLIED TO THE GLUCOSE-INSULIN METABOLISM

In this paper a method for model identification of biological systems described by stochastic linear differential equations using a new comp utational technique for statistical Bayesian inference, namely mixed g raphical models in the sense of Lauritzen and Wermuth, is presented. T he model is identified in terms of biological model parameters and noi se parameters. This non-linear estimation problem is solved by means o f an exact inference algorithm. The parameter estimates are given as a -posteriori distributions which can be interpreted as fuzzy possibilit y distributions. For model-based simulations of the underlying biologi cal system the model parameters are represented as uncertain parameter s with the distributions obtained from the estimation procedure, We ap ply the presented methods to a model for the glucose-insulin metabolis m: the Karlsburg model for type I diabetes. (C) 1998 Elsevier Science Ireland Ltd. All rights reserved.