regret
A measure of the difference between the optimal payoff and the payoff achieved by an AI agent in decision-making scenarios, often used to evaluate the performance of online learning algorithms and reinforcement learning agents.
- Efficient Kernelized Learning in Polyhedral Games beyond Full Information: From Colonel Blotto to Congestion Games
- Learning Equilibria from Data: Provably Efficient Multi-Agent Imitation Learning
- Near-Optimal Regret-Queue Length Tradeoff in Online Learning for Two-Sided Markets
- Near-Optimal Sample Complexity for Online Constrained MDPs
- Online Experimental Design With Estimation-Regret Trade-off Under Network Interference
- Robust Contextual Pricing
- Scalable Exploration via Ensemble++
- Tradeoffs between Mistakes and ERM Oracle Calls in Online and Transductive Online Learning
- True Impact of Cascade Length in Contextual Cascading Bandits