Search papers, labs, and topics across Lattice.
1
0
Q-learning can provably converge to optimal policies in non-Markovian environments, even with hard state aggregation, using a simple action-commitment strategy.
