A Bayesian Framework for Large-Scale Identification of Nonlinear Hybrid Systems

Abstract In this paper, a two-level Bayesian framework is proposed for the identification of nonlinear hybrid systems from large data sets by embedding it in four-stage procedure. At first stage, feature vector selection techniques are used to generate reduced-size set given training set. The resulting then identify system using method, where objective assign each point corresponding sub-mode model. third assignment train classifier separate original and determine all points. Finally, once every assigned sub-mode, estimator estimate regressor sub-system independently. method tested on three case studies.