ICML
2005
IEEE
14 years 5 months ago
2005
IEEE
The current framework of reinforcement learning is based on maximizing the expected returns based on scalar rewards. But in many real world situations, tradeoffs must be made amon...