摘要
双向聚类算法可以发现基因表达谱中隐藏的信息。为了寻找规模较大的基因相似矩阵,结合粒子群算法强大的搜索能力,提出了GP-Cluster双向聚类算法。基于粒子群(PSO)算法,引入Sigmoid函数进行动态调整,并在粒子飞行过程中加入了遗传算法(GA)"优胜劣汰"的思想,增加粒子运动的多变性和随机性,避免算法陷入局部最优。实验结果证明:相比GA算法和PSO算法,改进后的混合粒子群算法GP-Cluster能找到质量更佳的双向聚类,取得更好的聚类效果。
Bi-clustering algorithm can find the hidden information in the gene expression patterns.In order to find a larger genetic similarity matrix,combined with powerful search ability of particle swarm optimization,GP-Cluster algorithm was proposed.This algorithm was based on particle swarm algorithm(PSO),introducing the Sigmiod function to adjust the weights dynamically,and using the genetic algorithm(GA),the thought of "evolution ",to increase the variability and randomness of particle movements and to avoid algorithm falls into local optimum.The experimental results show that compared with GA algorithm and PSO algorithm,the improved hybrid particle swarm algorithm(GPCluster) can find a better quality of bi-clustering.
作者
李梁
陈佳瑜
LI Liang CHEN Jia-yu(College of Computer Science and Engineering, Chongqing University of Technology, Chongqing 400054, Chin)
出处
《重庆理工大学学报(自然科学)》
CAS
2017年第2期89-94,共6页
Journal of Chongqing University of Technology:Natural Science
基金
重庆市应用开发计划项目(CSTC2013yykf A40002)
关键词
数据挖掘
基因表达谱
双向聚类
粒子群算法
遗传算法
data mining
gene expression
bi-clustering
particle swarm algorithm
genetic algorithm