ECML
2005
Springer
15 years 8 months ago
2005
Springer
Abstract. While direct, model-free reinforcement learning often performs better than model-based approaches in practice, only the latter have yet supported theoretical guarantees f...