资源分类
》 资源 》论文》Contextual bandits with surrogate losses: Margin bounds and efficient algorithms

Contextual bandits with surrogate losses: Margin bounds and efficient algorithms

2020-02-17 | |  37 |   34 |   0

Abstract

 We use surrogate losses to obtain several new regret bounds and new algorithms for contextual bandit learning. Using the ramp loss, we derive new margin-based regret bounds in terms of standard sequential complexity measures of a benchmark class  of real-valued regression functions. Using the hinge loss, we derive an efficient algorithm with a image.png -type mistake bound against benchmark policies induced by d-dimensional regressors. Under realizability assumptions, our results also yield classical regret bounds.

上一篇：DeepExposure: Learning to Expose Photos with Asynchronously Reinforced Adversarial Learning

下一篇：The Lingering of Gradients: How to Reuse Gradients over Time

用户评价
全部评价

热门资源

  • Learning to Predi...

    Much of model-based reinforcement learning invo...

  • Stratified Strate...

    In this paper we introduce Stratified Strategy ...

  • The Variational S...

    Unlike traditional images which do not offer in...

  • A Mathematical Mo...

    Direct democracy, where each voter casts one vo...

  • Rating-Boosted La...

    The performance of a recommendation system reli...