中文版 | English
CSpace
An Improved Arithmetic Optimization Algorithm and Its Application to Determine the Parameters of Support Vector Machine
Fang, Heping1,2; Fu, Xiaopeng3; Zeng, Zhiyong4; Zhong, Kunhua1; Liu, Shuguang1
2022-08-01
摘要The arithmetic optimization algorithm (AOA) is a new metaheuristic algorithm inspired by arithmetic operators (addition, subtraction, multiplication, and division) to solve arithmetic problems. The algorithm is characterized by simple principles, fewer parameter settings, and easy implementation, and has been widely used in many fields. However, similar to other meta-heuristic algorithms, AOA suffers from shortcomings, such as slow convergence speed and an easy ability to fall into local optimum. To address the shortcomings of AOA, an improved arithmetic optimization algorithm (IAOA) is proposed. First, dynamic inertia weights are used to improve the algorithm's exploration and exploitation ability and speed up the algorithm's convergence speed; second, dynamic mutation probability coefficients and the triangular mutation strategy are introduced to improve the algorithm's ability to avoid local optimum. In order to verify the effectiveness and practicality of the algorithm in this paper, six benchmark test functions are selected for the optimization search test verification to verify the optimization search ability of IAOA; then, IAOA is used for the parameter optimization of support vector machines to verify the practical ability of IAOA. The experimental results show that IAOA has a strong global search capability, and the optimization-seeking capability is significantly improved, and it shows excellent performance in support vector machine parameter optimization.
关键词arithmetic optimization algorithm (AOA) dynamic inertia weights dynamic coefficient of mutation probability triangular mutation strategy support vector machine
DOI10.3390/math10162875
发表期刊MATHEMATICS
卷号10期号:16页码:20
通讯作者Fang, Heping(fangheping@cigit.ac.cn) ; Liu, Shuguang(liushuguang@cigit.ac.cn)
收录类别SCI
WOS记录号WOS:000845481200001
语种英语