OAS stochastic stochastic dynamic programming markov markov chains stochastic->markov method for finding numerically optimal policies in decisionunder decision making under uncertainty stochastic->decisionunder technique for solving problems of Bellman Bellman equation stochastic->Bellman represents the problem via a convexity convexity value value functions convexity->value property of objective objective function value->objective returns the value of Bellman->value describes the relationship between consecutive optimization optimization problem optimization->objective optimizes