| 基于复杂网络理论和遗传算法的分类方法 |
| |
| 引用本文： | 赵凤霞,谢福鼎,稽敏.基于复杂网络理论和遗传算法的分类方法[J].计算机应用与软件,2010,27(2):44-46. |
| |
| 作者姓名： | 赵凤霞 谢福鼎 稽敏 |
| |
| 作者单位： | 辽宁师范大学计算机与信息技术学院,辽宁,大连,116081 |
| |
| 基金项目： | 国家重点基础研究发展规划(2004CB318000) |
| |
| 摘 要： | 结合复杂网络研究的部分新理论成果和遗传算法的思想，提出一种新的分类方法。该方法将数据集按给定的相似度公式构造出具有社团结构的网络，在此网络的基础上用遗传算法的思想进行分类。算法引进社团模块度作为适应度函数，并且提出了节点归类错误率（NCM）对每次迭代产生的解进行纠错，提高了分类质量和速度。实验表明算法在分类精度和时间上都优于基于加权复杂网络特征的K—means聚类算法（WCNFC）。
|
| 关 键 词： | 复杂网络 遗传算法 社团模块度 节点归类错误率 |
A NEW CLASSIFICATION METHOD BASED ON COMPLEX NETWORK THEORY AND GENETIC ALGORITHM |
| |
| Zhao Fengxia,Xie Fuding,Ji Min.A NEW CLASSIFICATION METHOD BASED ON COMPLEX NETWORK THEORY AND GENETIC ALGORITHM[J].Computer Applications and Software,2010,27(2):44-46. |
| |
| Authors: | Zhao Fengxia Xie Fuding Ji Min |
| |
| Affiliation: | School of Computer and Information Technology/a>;Liaoning Normal University/a>;Dalian 116081/a>;Liaoning/a>;China |
| |
| Abstract: | A new classification method is proposed in combination with some new theory outcomes of the complex network research and the thought of genetic algorithm.The method converts data sets into a network with community structure according to given similarity formula firstly,and then based on the network it uses genetic algorithm thought to classify.The method introduces community modularity as the fitness formula,and proposes mistake rate of node classification(NCM) to clean up the solution of each iteration,thi... |
| |
| Keywords: | Complex network Genetic algorithm Community modularity Mistake rate of node classification |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
