G Hailu, G Sommer - ICANN 98: Proceedings of the 8th International …, 1998 - Springer
In almost all real systems where reinforcement learning is applied, it is found that a
knowledge free approach doesn't work. The basic RL algorithms must sufficiently be biased …