pipette
ESEspañol

Multiplicative Optimism for Constant Regret in Games

Ashkan Soleymani, Georgios Piliouras

Preprint

In the authors' words

We introduce Multiplicatively Optimistic Regret Matching (MORM), an uncoupled learning rule for finite general-sum games. Under simultaneous full-information self-play, every player achieves external regret uniformly over all horizons, using only one-step optimism. The analysis combines a potential-based regret-matching argument with multiplicative stability and Hellinger control of strategy movement. A learning-rate safeguard additionally gives regret in the face of adversarial utilities.

Main resultLimitation the authors admit

Appeared: Monday, September 21. arXiv. Preprint, not yet peer-reviewed.