Computer Engineering and Applications ›› 2021, Vol. 57 ›› Issue (16): 50-64.DOI: 10.3778/j.issn.1002-8331.2104-0096

Previous Articles     Next Articles

Research Progress on Parameter Optimization of SVR Based on Swarm Intelligence Algorithm

ZHANG Lin, WANG Tinghua, ZHOU Huiying   

  1. School of Mathematics and Computer Science, Gannan Normal University, Ganzhou, Jiangxi 341000, China
  • Online:2021-08-15 Published:2021-08-16



  1. 赣南师范大学 数学与计算机科学学院,江西 赣州 341000


Support Vector Regression(SVR) is a machine learning method based on statistical learning theory, and is mainly used to deal with regression problems. Selecting appropriate parameters is a prerequisite for realizing the advantages of the SVR algorithm, but in practice there is still a difficulty in selecting model parameters. The swarm intelligence algorithm is a meta-heuristic algorithm that imitates the social behavior of biological populations in the natural world. It has the characteristics of simplicity, adaptability and flexibility, and has become a research hotspot of nonlinear parameter optimization methods. The research progress using swarm intelligence algorithm to optimize the parameters of SVR is comprehensively reviewed. Specifically, after introducing the basic theory of SVR, it systematically analyzes the advantages and shortcomings of common swarm intelligence algorithms and their improved methods for realizing the selection and optimization of SVR parameters, and finally gives prospects for future research directions and challenges.

Key words: Support Vector Regression(SVR), parameter optimization, swarm intelligence algorithm, machine learning


支持向量回归机(Support Vector Regression,SVR)是建立在统计学习理论上的一种机器学习方法,主要用来处理回归问题。选取到合适的参数是实现支持向量回归机算法优势的前提,但在实践中仍然存在模型参数选择困难的问题。群智能算法主要是模仿自然界生物种群社会行为规律的元启发式算法,具有简单性、自适应性、灵活性等特点,现已成为非线性参数寻优方法的研究热点。系统综述了利用群智能算法优化支持向量回归机参数的研究进展。在介绍支持向量回归机基础理论之后,系统分析了常见群智能算法及其改进方法实现支持向量回归机参数优化选择的优点与不足,并对未来的研究方向及挑战做出展望。

关键词: 支持向量回归机(SVR), 参数优化, 群智能算法, 机器学习