基于打分准则和改进PSO的基因选择方法  

Gene selection method based on gene scoring strategy and improved particle swarm optimization

作　　者：唐迪[1] 韩飞[1] 程准[2] 

机构地区：[1]江苏大学计算机科学与通信工程学院,江苏镇江212013 [2]南京农业大学工学院,江苏南京210031

出　　处：《计算机工程与设计》2018年第3期710-715,共6页Computer Engineering and Design

基　　金：国家自然科学基金项目(61572241)

摘　　要：为利用数据分析的方法高效快速筛选出具有高分类性能的基因,针对基因表达谱数据高维小样本的特点,提出一种基于打分准则和改进的PSO算法的基因选择方法。基于数学抽样调查的科学性,制定一种基因打分准则来准确筛选相关基因;为防止粒子陷入局部最优解,利用半初始化及模拟退火算法的Metropolis准则改进PSO算法。在两个公开的数据集上的实验结果表明,该方法快捷、高效,克服了传统聚类解释性差以及PSO算法易于陷入局部最优解的缺点,选出了数目较少分类性能较强的基因。To obtain genes highly related to sample＇s class with the method of data analysis rapidly and effectively,aiming at the characteristics of gene expression data with small sample in comparison to high dimensionality,a hybrid gene selection method based on gene scoring strategy and improved particle swarm optimization（PSO）was proposed.Based on the scientific nature of mathematical sampling,the gene scoring strategy was proposed to filter irrelevant genes.To prevent the swarm from trapping into local minima,the particle swarm optimization（PSO）was improved using half-initialization and metropolis criterion of the simulated annealing algorithm（SA）.The experiments on two public datasets show that the proposed method is fast and efficient.It can overcome the shortcomings that the traditional clustering is poor explicable and the PSO algorithm is easy to fall into local optimal solution.The proposed method can select out the effective gene set with fewer number and good classification performance.

关 键 词：基因选择 基因打分准则 半初始化 METROPOLIS准则 微粒群算法 

分 类 号：Q811.4[生物学—生物工程] TP18[自动化与计算机技术—控制理论与控制工程]