ICML
2004
IEEE
15 years 7 days ago
2004
IEEE
We consider learning in a Markov decision process where we are not explicitly given a reward function, but where instead we can observe an expert demonstrating the task that we wa...