Kernelized value function approximation for reinforcement learning.

Gavin Taylor, Ronald Parr

Browse the full ICML paper archive.