Multiplicative updates outperform generic no-regret learning in congestion games. (2009)

by R Kleinberg, G Piliouras, E Tardos
Venue:In Proc. 40th ACM Symposium on Theory of Computing,