We describe an extension to the Mixture of Experts architecture for modelling and controlling dynamical systems which exhibit multiple modesof behavior. This extension is based on a Markov process model, and suggests a recurrent network for gating a set of linear or non-linear controllers. The new architecture is demonstrated to be capable of learning e ective control strategies for jump linear and non-linear plants with multiple modes of behavior.
Timothy W. Cacciatore, Steven J. Nowlan