91
Voted
ICML
2003
IEEE
16 years 11 days ago
2003
IEEE
We present a learning framework for Markovian decision processes that is based on optimization in the policy space. Instead of using relatively slow gradient-based optimization al...