Bootstrapping of parameterized skills through hybrid optimization in task and policy spaces

reinforcement learning, policy optimization, memory, learning, hybrid optimization, dimensionality reduction, parameterized skills, ddc:000