摘要
提出了函数挖掘成功率、弱相关和函数一致性合并的概念,在此基础上给出了基于网格的GEP函数挖掘算法(GEPFM-grid,gene expression programming function mining based upon grid)。通过比较实验表明,GEPFM-grid的函数挖掘成功率和收敛速度比传统算法有着明显的提升且耗时较少。
The concepts of success rate of function mining, weak correlation and merger of function consistency were proposed. On the basis of these, gene expression programming function mining based on grid （GEPFM-Grid） was put forward. By extensive experiment of GEPFM-grid and other traditional algorithms, the results show that success rate of function mining and convergent speed of GEPFM-grid is obviously improved, and that consumptive time is less.
出处
《通信学报》
EI
CSCD
北大核心
2008年第6期69-74,共6页
Journal on Communications
关键词
函数挖掘
基因表达式编程
网格
弱相关
function mining
gene expression programming
grid
weak correlation
