Reinforcement learning algorithms that assimilate and accommodate skills with multiple tasks | IEEE Conference Publication | IEEE Xplore