摘要
针对灰狼优化算法收敛速度慢、寻优精度低、易陷入局部最优等缺陷,提出一种基于差分进化(DE)的灰狼优化算法(GWODE).该算法在灰狼优化算法的基础上,引进差分进化机制生成变异种群,通过调节缩放因子和交叉概率因子避免算法陷入局部最优.引入精英保留策略,根据进化后狼群适应度进行排序,淘汰适应度差的灰狼,同时再引进相同数量灰狼确保种群的竞争力.本文将该算法应用于生物医学诊断方面.实验结果表明,本文提出的算法性能优于实验对比的特征选择算法.
In real application,Grey Wolf Optimizer may encounter lots of problems,such as slow convergence speed,low convergence precision and easy to fall into local optimal solution.To avoid these problems,we propose a novel method named Grey Wolf Optimizer based on Differential Evolution(GWODE).In GWODE,differential evolution is introduced to produce population variation,and the adjustment of scale factor and cross probability factor is adopted to avoid local optimal solution.To pursue higher competitiveness,the wolf of low fitness value will be eliminated,and the same number of new wolves will be brought.Finally,we introduce this method into biomedical diagnosis application.Experiments results show that our method gets a better performance than the compared methods.
作者
王俊
冯军
张戈
王建林
王胜
郑泰皓
WANG Jun;FENG Jun;ZHANG Ge;WANG Jianlin;WANG Sheng;ZHENG Taihao(School of Computer and Information Engineering,Henan University,Henan Kaifeng475004,China)
出处
《河南大学学报（自然科学版）》
CAS
2020年第5期570-578,共9页
Journal of Henan University:Natural Science
基金
国家自然科学基金资助项目(61802114,61802113)
河南省高等院校重点科研项目(18A120001,18A520021)。
关键词
灰狼优化算法
差分进化
生物医学诊断
特征选择
gray wolf optimization algorithm
differential evolution
biomedical diagnosis
feature selection
