针对基本功能流聚类算法计算复杂度高、聚类正确率较低的缺点,提出一种基于改进的果蝇算法与功能流算法相融合的聚类分析算法Flow-IFOA。通过引入果蝇因子,根据离最优解果蝇的距离自适应地调整每个果蝇个体的搜索步长,保证了算法的搜索精度和速度。将改进后的果蝇算法与功能流算法融合,在PPI网络数据库上的仿真结果表明,改进算法相比其他聚类算法得到了较好的聚类正确率和较快的收敛速度,是一种行之有效的方法。%The paper proposes a new clustering analysis algorithm FLOW-IFOA, which is based on the combination of improved fruit fly algorithm and function flow clustering algorithm and aims at the drawbacks of basic function flow clustering algorithm such as high time com-plexity and low clustering accuracy.The searching accuracy and speed of the proposed algorithm are guaranteed by introducing the fruit fly factor and adaptively adjusting the searching step size of individual fruit fly according to the distance to optimal solution fruit fly.Simulations on the PPI network database indicate that the proposed algorithm has better clustering accuracy and faster converging rate compared with other clustering algorithms, so it is an effective method.
展开▼