Minimizing Dynamic Regret and Adaptive Regret Simultaneously.
Lijun ZhangShiyin LuTianbao YangPublished in: CoRR (2020)
Keyphrases
- online learning
- confidence bounds
- lower bound
- expert advice
- regret minimization
- dynamically adjust
- loss function
- minimax regret
- worst case
- binary classification
- information systems
- bandit problems
- weighted majority
- upper confidence bound
- reward function
- regret bounds
- data driven
- multi class
- active learning
- decision trees
- decision making
- learning algorithm
- online convex optimization