This work introduces scheduling strategies to maximize the expected numberof independent tasks that can be executed on a cloud platform within a given budgetand under a deadline constraint. Task execution times are not known before execution;instead, the only information available to the scheduler is that they obey some (unknown)probability distribution. The scheduler needs to acquire some information before decidingfor a cutting threshold: instead of allowing all tasks to run until completion, one maywant to interrupt long-running tasks at some point. In addition, the cutting thresholdmay be reevaluated as new information is acquired when the execution progresses further.This works presents several strategies to determine a good cutting th...