Contextual Bandits with Continuous Actions: Smoothing, Zooming, and Adapting.
Akshay KrishnamurthyJohn LangfordAleksandrs SlivkinsChicheng ZhangPublished in: CoRR (2019)
Keyphrases
- action space
- context sensitive
- contextual information
- continuous action
- plan recognition
- stochastic systems
- decision theoretic
- reasoning about actions
- image smoothing
- neural network
- state space
- spatio temporal
- context dependent
- computer vision
- real time
- goal directed
- multiple agents
- image filtering
- search engine
- smoothing methods
- smoothing algorithm
- learning algorithm