Approximate dynamic programming (ADP) is the standard technique to derive optimal policies in finite-horizon stochastic multistage optimal decision problems, with continuous state space. Yet, it ...