摘要
基因表达式程序设计(GEP)是一种基于基因型和表现型的新型进化算法。针对传统GEP遗传算子进行建模及预测时容易受到噪声干扰,导致过早收敛,陷入局部最优等缺点,提出了一种新型GEP算法,增加了"倒串"和"基因提取"算子,该新型算法可以提高基因的有效利用率,具有更高的收敛速度和求解精度,且能更好的避免早熟现象。将该新型GEP算法用在菠菜价格预测上,通过对训练数据进行分析和进化创建数学模型,实现菠菜价格的仿真与预测,通过多组实验证明,该新型基因表达式程序设计算法在菠菜价格预测上具有更快的收敛速度和更高的精度。
Gene expression programming（GEP） is a new evolutionary algorithm based on genotype and phenotype, the genetic operators of traditional GEP are easily influenced by noise, leading to premature convergence and local solution. This paper proposes an improved GEP（IGEP）algorithm, which increases the ＂inverted series＂ and ＂extract＂ operator, can effectively increase the utilization rate of genes, and the convergence speed and solution precision is higher, can avoid the premature phenomenon. The improved GEP algorithm is used to predict the price of spinach. Through the analysis and evolution of training data to create a mathematical model,the simulation and prediction of spinach price is realized. The results show that the new IGEP algorithm has faster convergence speed and higher accuracy in the prediction of spinach price.
出处
《广东农业科学》
CAS
2016年第1期151-158,共8页
Guangdong Agricultural Sciences
基金
广东省科技计划项目(2015A020209119
2014A020208087
2012B040500054)
关键词
GEP
IGEP
菠菜价格预测
基因利用率
基因提取
gene expression programming
improved gene expression programming
spinach price prediction
utilization of gene
gene extraction
