
软件学报 ISSN 1000-9825, CODEN RUXUEW E-mail: jos@iscas.ac.cn
Journal of Software,2018,29(10):30683090 [doi: 10.13328/j.cnki.jos.005607] http://www.jos.org.cn
©中国科学院软件研究所版权所有. Tel: +86-10-62562563
贝叶斯优化方法和应用综述
崔佳旭
1,2
,
杨
博
1,2
1
(符号计算与知识工程教育部重点实验室(吉林大学),吉林 长春 130012)
2
(吉林大学 计算机科学与技术学院,吉林 长春 130012)
通讯作者: 杨博, E-mail: ybo@jlu.edu.cn
摘 要: 设计类问题在科学研究和工业领域无处不在.作为一种十分有效的全局优化算法,近年来,贝叶斯优化方
法在设计类问题上被广泛应用.通过设计恰当的概率代理模型和采集函数,贝叶斯优化框架只需经过少数次目标函
数评估即可获得理想解,非常适用于求解目标函数表达式未知、非凸、多峰和评估代价高昂的复杂优化问题.从方
法论和应用领域两方面深入分析、讨论和展望了贝叶斯优化的研究现状、面临的问题和应用领域,期望为相关领
域的研究者提供有益的借鉴和参考.
关键词: 贝叶斯优化;全局优化算法;概率代理模型;采集函数;黑箱
中图法分类号: TP18
中文引用格式: 崔佳旭,杨博.贝叶斯优化方法和应用综述.软件学报,2018,29(10):30683090. http://www.jos.org.cn/1000-
9825/5607.htm
英文引用格式: Cui JX, Yang B. Survey on Bayesian optimization methodology and applications. Ruan Jian Xue Bao/Journal of
Software, 2018,29(10):30683090 (in Chinese). http://www.jos.org.cn/1000-9825/5607.htm
Survey on Bayesian Opti mization Methodology a nd Applications
CUI Jia-Xu
1,2
, YANG Bo
1,2
1
(Key Laboratory of Symbolic Computation and Knowledge Engineering for the Ministry of Education (Jilin University), Changchun
130012, China)
2
(College of Computer Science and Technology, Jilin University, Changchun 130012, China)
Abstra ct : Designing problems are ubiquitous in science research and industry applications. In recent years, Bayesian optimization,
which acts as a very effective global optimization algorithm, has been widely applied in designing problems. By structuring the
probabilistic surrogate model and the acquisition function appropriately, Bayesian optimization framework can guarantee to obtain the
optimal solution under a few numbers of function evaluations, thus it is very suitable to solve the extremely complex optimization
problems in which their objective functions could not be expressed, or the functions are non-convex, multimodal and computational
expensive. This paper provides a detailed analysis on Bayesian optimization in methodology and application areas, and discusses its
research status and the problems in future researches. This work is hopefully beneficial to the researchers from the related communities.
Key words: Bayesian optimization; global optimization algorithm; probabilistic surrogate model; acquisition function; black-box
设计类问题在科学研究和工业设计等领域无处不在.例如:编程人员通过选择恰当的算法来优化系统性能;
环境学家通过设计传感器部署位置来监控环境状况;化学家通过设计实验来获取新的物质;制药厂商通过设计
新型药物来抵抗疾病;食品厂商通过设计新的食谱来生产优质食品等等.通常,将这些设计问题考虑成如下最优
基金项目: 国家自然科学基金(61572226, 61876069); 吉林省重点科技研发项目(20180201067GX, 20180201044GX)
Foundation item: National Natural Science Foundation of China (61572226, 61876069); Jilin Province Key Scientific and
Technological Research and Development Project (20180201067GX, 20180201044GX)
收稿时间: 2017-06-12; 修改时间: 2018-04-02; 采用时间: 2018-05-17; jos 在线出版时间: 2018-06-07
CNKI 网络优先出版: 2018-06-07 14:53:49, http://kns.cnki.net/kcms/detail/11.2560.TP.20180607.1453.008.html
评论