arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2310.02767cs.LG
收藏

动态与非平稳环境中的基于核的函数学习

Kernel-based function learning in dynamic and non stationary environments

University of Cambridge(剑桥大学);University of Padova(帕多瓦大学)

Alberto Giaretta, Mauro Bisiacco, Gianluigi Pillonetto

AI总结:

针对非平稳时变分布下的函数估计问题,本文研究基于核的岭回归,推导其收敛条件,并覆盖随机适应无限频繁发生及多智能体探索-利用场景。

摘要:

机器学习的一个核心主题是从稀疏且含噪的数据中进行函数估计。一个例子是监督学习,其中训练集的元素是数据对,每一对都包含一个输入位置和一个输出响应。在过去几十年中,大量研究致力于设计未知函数的估计量,并研究它们向最优预测器的收敛性,同时刻画学习率。这些结果通常依赖于平稳性假设,即输入位置从一个不随时间变化的概率分布中抽取。在本工作中,我们考虑基于核的岭回归,并在非平稳分布下推导收敛条件,同时处理随机适应可能无限频繁发生的情形。这包括重要的探索-利用问题,例如一组智能体/机器人必须监测环境以重建一个传感场,并且它们的运动规则会基于已获得的关于该场和/或周围环境的知识而持续更新。

英文摘要:

One central theme in machine learning is function estimation from sparse and noisy data. An example is supervised learning where the elements of the training set are couples, each containing an input location and an output response. In the last decades, a substantial amount of work has been devoted to design estimators for the unknown function and to study their convergence to the optimal predictor, also characterizing the learning rate. These results typically rely on stationary assumptions where input locations are drawn from a probability distribution that does not change in time. In this work, we consider kernel-based ridge regression and derive convergence conditions under non stationary distributions, addressing also cases where stochastic adaption may happen infinitely often. This includes the important exploration-exploitation problems where e.g. a set of agents/robots has to monitor an environment to reconstruct a sensorial field and their movements rules are continuously updated on the basis of the acquired knowledge on the field and/or the surrounding environment.

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。

↑