Automata behavior analysis in stationary and composite random medium

This Master’s thesis discusses the special class of finite stochastic machine that operates in a stationary stochastic environment, which reacts with the operation of the automatic response and it's perceived as an automatic reaction - one of the types of reactions in the following three classes of reaction: the desired reaction class (profits, rewarding), Class of adverse reactions (loss, fine) and class-neutral reactions (indifference). The operation of the stochastic machine’s describing Markov chains eigenvalues importance of the task is decided in this environment. Based on the method of this class of machine productions functions studied the behavior of some of the asymptotic behavior of the machines possible complete classification; It is also studied the behavior in a composite environment, which consists of stationary circumstances and their switching is performed by Markov chains.