Models of social phenomena and scenarios approaches in decision-making

Abstract

A new class of models is proposed with the properties of associative memory to be used in investigation of large social systems. The models have the structure similar to neuronets models of the Hopfield type. Accounting the intellectual properties allowed extending the scope of modelable phenomena. In particular, the individual ability for forecasting and decision-making results in new properties of solutions. The main new property is possible appearance of multivalued solutions. For large social systems, this leads to the existence of many trajectories of system, that is, many scenarios of behavior. Some analogies with quantum-mechanical phenomena are discussed.