Timezone: »

Online Convex Optimization with Stochastic Constraints
Hao Yu · Michael Neely · Xiaohan Wei

Wed Dec 06 06:30 PM -- 10:30 PM (PST) @ Pacific Ballroom #55 #None
This paper considers online convex optimization (OCO) with stochastic constraints, which generalizes Zinkevich's OCO over a known simple fixed set by introducing multiple stochastic functional constraints that are i.i.d. generated at each round and are disclosed to the decision maker only after the decision is made. This formulation arises naturally when decisions are restricted by stochastic environments or deterministic environments with noisy observations. It also includes many important problems as special case, such as OCO with long term constraints, stochastic constrained convex optimization, and deterministic constrained convex optimization. To solve this problem, this paper proposes a new algorithm that achieves $O(\sqrt{T})$ expected regret and constraint violations and $O(\sqrt{T}\log(T))$ high probability regret and constraint violations. Experiments on a real-world data center scheduling problem further verify the performance of the new algorithm.

Author Information

Hao Yu (University of Southern California)
Michael Neely (Univ. Southern California)
Xiaohan Wei (University of Southern California)

More from the Same Authors