English

On the asymptotic optimality of the comb strategy for prediction with expert advice

Probability 2020-03-23 v2 Machine Learning Optimization and Control

Abstract

For the problem of prediction with expert advice in the adversarial setting with geometric stopping, we compute the exact leading order expansion for the long time behavior of the value function. Then, we use this expansion to prove that as conjectured in Gravin et al. [12], the comb strategies are indeed asymptotically optimal for the adversary in the case of 4 experts.

Keywords

Cite

@article{arxiv.1902.02368,
  title  = {On the asymptotic optimality of the comb strategy for prediction with expert advice},
  author = {Erhan Bayraktar and Ibrahim Ekren and Yili Zhang},
  journal= {arXiv preprint arXiv:1902.02368},
  year   = {2020}
}

Comments

To appear in Annals of Applied Probabilty