NIPS
2000
14 years 1 months ago
2000
We present an expressive agent design language for reinforcement learning that allows the user to constrain the policies considered by the learning process.The language includes s...