A Sharp Memory-Regret Trade-Off for Multi-Pass Streaming Bandits.
Arpit AgarwalSanjeev KhannaPrathamesh PatilPublished in: CoRR (2022)
Keyphrases
- trade off
- online learning
- multi armed bandit problems
- regret bounds
- data streams
- lower bound
- multi armed bandit
- real time
- multi armed bandits
- limited memory
- memory usage
- neural network
- least squares
- memory requirements
- streaming data
- expert advice
- database
- multi class
- main memory
- pairwise
- computing power
- data structure
- high quality
- bandit problems
- bias variance
- data sets