Title :
Decision making for a great number of alternatives by self-organisation
Author_Institution :
Central Sci. & Res. Inst., State Res. Center of Russia, Saint-Petersburg, Russia
Abstract :
In this paper we consider the of statistical decisions problem for the alternatives formed by different system states and their combinations. Decisions based on the measurements are made by a posteriori probabilities of the alternatives. The analogy is shown between the differential equations for a posteriori probabilities and Eigen´s model for the evolution of biological macromolecules. Relying on this analogy it is shown that at a posteriori probability updating the alternatives always compete and only the best alternative survive. Rather than use for decision making the model with competitive behavior in this paper it is proposed the use of Epstein´s model for the evolution. This model has the property of self-organization because it allows both competitive and cooperative behavior, as the situation requires. The parametrical control has been suggested for the equations behavior in the interests of test of hypothesis.
Keywords :
decision making; differential equations; evolution (biological); macromolecules; molecular biophysics; probability; self-adjusting systems; statistical analysis; Eigen model; Epstein model; biological macromolecular evolution; differential equations; hypothesis test; parametrical control; posteriori probabilities; self-organisation; statistical decision making; system states; Biological system modeling; Condition monitoring; Decision making; Differential equations; Evolution (biology); Molecular biophysics; Nonlinear equations; Probability; Testing; Vectors;
Conference_Titel :
Physics and Control, 2005. Proceedings. 2005 International Conference
Print_ISBN :
0-7803-9235-3
DOI :
10.1109/PHYCON.2005.1513951