Controlled Markov processes, Gittins index, Long run average return, Markov decision process, Multiarmed bandit problem, Multicomponent systems, No discounting, Optimal strategy, Simple family of alternative Markov bandit processes, Switching penalties
