Reducing policy degradation in neuro-dynamic programming.

Thomas Gabel, Martin A. Riedmiller

Browse the full ESANN paper archive.