论文部分内容阅读
将快速退火演化算法(fastannealingevolutionaryalgorithm,FAEA)与协同方法相结合,提出了一种用于求解高维的全局优化问题的新方法——协同快速退火演化算法(cooperativefastannealingcoevolutionaryalgorithm,CFACA).首先将高维的解空间分解成多个一维的子空间,再在每个子空间里利用单个独立的FAEA搜索该子空间里的最优子解,最后将各子解结合在一起,即构成了原来问题的一个解.基准函数测试的结果表明,CFACA算法具有更快的收敛速度.进一步用CFACA算法提取EGF蛋白质家族的模体,正确识别率达到67.0%,所提取的模体与蛋白质功能位点数据库PROSITE中的结果相吻合.
Combining the fast annealing evolutionary algorithm (FAEA) and the collaborative method, a new method for solving the global optimization problem with high dimensions, called cooperative fastancaling evolutionary algorithm (CFACA), is proposed.Firstly, the high dimensional The solution space is decomposed into a plurality of one-dimensional subspaces. In each subspace, a single independent FAEA is used to search for the optimal sub-solution in the subspace. Finally, the sub-solutions are combined to form one of the original problems The results of the benchmark function test showed that the CFACA algorithm has a faster convergence rate.Furthermore, the correct recognition rate of 67% was obtained by extracting the motif of the EGF protein family using the CFACA algorithm. The extracted motifs and protein functional site database PROSITE The results match.