An Off-Policy Natural Policy Gradient Method for a Partial Observable Markov Decision Process.

Yutaka Nakamura, Takeshi Mori, Shin Ishii

Browse the full ICANN paper archive.