Movement Penalized Bayesian Optimization with Application to Wind Energy Systems

In this setting, the learner receives context (e.g., weather conditions) at each round, and has to choose an action (e.g., turbine parameters).