Approximate solutions to convex optimization problems under stochastic uncertainty