Mathematical theory of discrete time decision processes, also known as stochastic control, is based on two major ideas: backward induction and conditioning. We cover systems with complete or partial information as well as with complete or partial observation.