TY  - RPRT
TI  - A unified analysis of value-function-based reinforcement- learning algorithms.
AU  - C Szepesvári
AU  - M L Littman
PY  - 1999
DO  - 10.1162/089976699300016070
UR  - https://pubmed.ncbi.nlm.nih.gov/10578043/
ID  - 10578043
ER  -