TY  - RPRT
TI  - Biased Aggregation, Rollout, and Enhanced Policy Improvement for Reinforcement Learning
AU  - Dimitri Bertsekas
PY  - 2019
UR  - https://arxiv.org/abs/1910.02426
ID  - 1910.02426
ER  -