摘要
提出了一种新的基于时间序列分析的GEP数据预测方法,通过向基因表达式编程(GEP)中引入突变判断和回溯技术,提高了拟合成功率和函数适应度,避免“早熟”问题发生。最后利用编程和实验对新方法的效果进行了验证。
A new approach for data prediction based on time series prediction is proposed. By incorporating mutation estimate and backtracking in gene expression programming, the rate of fitting and the fit degree are increased, and the problem of premature is avoided. Finally, programme and experiment are done to confirm the effect.
出处
《仪器仪表学报》
EI
CAS
CSCD
北大核心
2006年第z2期1394-1395,共2页
Chinese Journal of Scientific Instrument
关键词
基因表达式编程
时间序列分析
回溯
突变判断
gene expression programming time series prediction backtracking mutation estimate
