CORR
2010
Springer
13 years 3 months ago
2010
Springer
We consider model-based reinforcement learning in finite Markov Decision Processes (MDPs), focussing on so-called optimistic strategies. Optimism is usually implemented by carryin...