Iterative Value-Aware Model Learning

V alue-A ware Model Learning (V AML) is a novel framework for learning the model of the environment in Model-Based Reinforcement Learning (MBRL) [Farahmand et al., 2017a, 2016a].