ICML
2008
IEEE
14 years 5 months ago
2008
IEEE
We consider the task of reinforcement learning in an environment in which rare significant events occur independently of the actions selected by the controlling agent. If these ev...