//-->
Value iteration and approximately optimal stationary policies in finite-state average Markov decision chains
Cavazos-Cadena, Rolando, (2002)