111
click to vote
AAAI
1996
15 years 4 months ago
1996
We formalize a model for supervised learning of action strategies in dynamic stochastic domains and show that PAC-learning results on Occam algorithms hold in this model as well. W...