Online Optimization

Follow the Regularized Leader (FTRL)

Variant of FTL incorporating regularization to stabilize sequential decisions and guarantee better regret bounds in adversarial environments.

← Tillbaka