arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2007.05670stat.MLcs.LGmath.STstat.TH
收藏

渐近最优多臂Bandit算法及超参数优化

An Asymptotically Optimal Multi-Armed Bandit Algorithm and Hyperparameter Optimization

Huawei Noah’s Ark Lab(华为诺亚方舟实验室);Shanghai Jiao Tong University(上海交通大学);Peking University(北京大学)

Yimin Huang, Yujun Li, Hanrong Ye, Zhenguo Li, Zhihua Zhang

摘要:

在具有大超参数搜索空间的高级深度学习中,超参数、神经架构或数据增强策略的评估成为关键的模型选择问题。在本文中,我们提出了一种有效的和强大的基于带宽的算法称为子采样(SS)的超参数搜索评估的情况下。该算法通过观测值的子样本来评估超参数的潜力,并在理论上证明了其在累积后悔准则下的最优性。我们进一步结合SS和贝叶斯优化,提出了一种新的超参数优化算法BOSS。实证研究验证了我们的SS理论论点,并证明了BOSS在许多应用程序上的卓越性能,包括神经架构搜索(NAS),数据增强(DA),对象检测(OD)和强化学习(RL)。

英文摘要:

The evaluation of hyperparameters, neural architectures, or data augmentation policies becomes a critical model selection problem in advanced deep learning with a large hyperparameter search space. In this paper, we propose an efficient and robust bandit-based algorithm called Sub-Sampling (SS) in the scenario of hyperparameter search evaluation. It evaluates the potential of hyperparameters by the sub-samples of observations and is theoretically proved to be optimal under the criterion of cumulative regret. We further combine SS with Bayesian Optimization and develop a novel hyperparameter optimization algorithm called BOSS. Empirical studies validate our theoretical arguments of SS and demonstrate the superior performance of BOSS on a number of applications, including Neural Architecture Search (NAS), Data Augmentation (DA), Object Detection (OD), and Reinforcement Learning (RL).

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑