Online Network Revenue Management Using Thompson Sampling

Online Network Revenue Management using Thompson Sampling

Thompson Sampling for Complex Online Problems

We consider stochastic multi-armed bandit problems with complex actions over a set of basic arms, where the decision maker plays a complex action rather than a basic arm in each round. The reward of the complex action is some function of the basic arms’ rewards, and the feedback observed may not necessarily be the reward perarm. For instance, when the complex actions are subsets of the arms, we...

Thompson Sampling for Complex Online Problems

We study stochastic multi-armed bandit settings with complex actions over a set ofbasic arms, where the decision maker has to select a subset of the basic arms or apartition of the basic arms at every round (rather than only selecting a single basicarm). The reward of the complex action is some function of the basic arms’ re-wards, and the feedback observed may not necessarily b...

Thompson sampling with the online bootstrap

Thompson sampling provides a solution to bandit problems in which new observations are allocated to arms with the posterior probability that an arm is optimal. While sometimes easy to implement and asymptotically optimal, Thompson sampling can be computationally demanding in large scale bandit problems, and its performance is dependent on the model fit to the observed data. We introduce bootstr...

Online Companion for Robust Controls for Network Revenue Management

Proof of Proposition 1. With a slight abuse of notation, let us denote by D d the set of multivariate stochastic processes such that the aggregate demand equals d. Accordingly, Problem (5) can be decomposed as follows: ρ(y) = max d∈P max D∈D d max z∈F R(z, D) − R(y, D) = max d∈P max z∈F max D∈D d R(z, D) + max D∈D d −R(y, D). Let ξ j be the realized sales under a booking policy z when the deman...