中文

带随机噪声的在线广义线性回归及其在异方差赌博机中的应用

机器学习 2023-03-28 v2 最优化与控制 机器学习

摘要

我们研究随机设定下的在线广义线性回归问题,其中标签由带有可能无界加性噪声的广义线性模型生成。我们对经典的跟随正则化领导者(FTRL)算法进行了尖锐分析以应对标签噪声。更具体地,对于σ\sigma-次高斯标签噪声,我们的分析给出了O(σ2dlogT)+o(logT)O(\sigma^2 d \log T) + o(\log T)的悔界上界,其中dd为输入向量维度,TT为总轮数。我们还证明了随机在线线性回归的Ω(σ2dlog(T/d))\Omega(\sigma^2d\log(T/d))下界,表明我们的上界近乎最优。此外,我们将分析推广到更精细的Bernstein噪声条件。作为应用,我们研究了带异方差噪声的广义线性赌博机,并基于FTRL提出一种算法以实现首个感知方差的悔界。

关键词

引用

@article{arxiv.2202.13603,
  title  = {Optimal Online Generalized Linear Regression with Stochastic Noise and Its Application to Heteroscedastic Bandits},
  author = {Heyang Zhao and Dongruo Zhou and Jiafan He and Quanquan Gu},
  journal= {arXiv preprint arXiv:2202.13603},
  year   = {2023}
}

备注

27 pages, 3 figures. In this updated version, we have changed the paper title, added new theoretical results on the FTRL algorithm and mainly focused on stochastic online regression. Refer to arXiv:2202.13603v1 for the previous version, which contains more results on heteroscedastic nonlinear bandits