|
|
Close Help | ||||||||||||||
The project is concerned with identification and control of uncertain systems, using Bayesian decision-making theory. The main advantage of this theory is consistency of the generated decision (i.e. estimates and control actions). However, solution of the implied recursive Bayesian relations is often available only in approximate form. An extension of the Bayesian theory for multiple decision makers (i.e. decentralized control) is studied at the department of adaptive systems, ÚTIA AV ČR. The proposed solution defines merging as a new probabilistic operation which has to be approximated for complex models.
Software toolbox mixtools 3000 is being developed as a platform implementing full process of decision making for testing of algorithms implied by the new theory. At present, only analytically tractable models (e.g. linear Gaussian) are supported. Implementation of all operations for static mixtures these models is planned.
Sampling methods are traditional approximation methodology of Bayesian statistics. Any complex probability density function can be approximated by a set of samples generated from it. This method is computationally expensive, however research effort to increase efficiency of sampling methods and increasing performance of computers improved applicability of these methods in such a way that they bring significant improvement in many application areas and represent a strong alternative to traditional approximation methods.
