蚁群聚类算法在协作学习分组中的应用研究  

Study on application of ant colony clustering algorithm in the grouping of collaborative learning

作　　者：张颀 ZHANG Qi(Wuyi University,Wuyishan 354300,China)

机构地区：[1]武夷学院信息技术与实验室管理中心,福建武夷山354300

出　　处：《新余学院学报》2021年第2期21-28,共8页Journal of Xinyu University

基　　金：福建省教育厅科技项目“多Agent网络协作学习模型的设计与实现”(JAT160509)。

摘　　要：为了改善传统的协作学习分组中存在的方式单一、缺乏智能性等缺点,将蚁群聚类算法应用于协作学习分组,并以大学计算机课程为例进行了算法验证。算法通过多个学生特征参数计算相似性,提高了聚类的准确性;合并一部分较为相似的学生数据,减少了实验数据,提高聚类效率;再通过计算数据领域内的平均相似性,得到聚类结果;最后在每个聚类中选择数据组成协作小组。实验结果表明,与传统协作学习相比,蚁群聚类算法的分组效率更高,分组结果更加合理有效,协作学习效果也得到明显改善。In order to improve the singleness of mode and lack of intelligence in the grouping of traditional collaborative learning,the ant colony clustering algorithm is applied to the grouping of collaborative learning,and the algorithm is verified through the grouping application of university computer courses.The similarity is calculated through multiple characteristic parameters of students,so as to improve the accuracy of clustering;some similar data of students merge together,so as to reduce experimental data and improve clustering efficiency;then the average similarity in adjacent area of data is calculated to get the clustering results;the data in each cluster are selected to form a collaborative group finally.The results of experiment show that the ant colony clustering algorithm has higher efficiency in grouping,and the grouping results are more reasonable and effective,and the effectiveness of collaborative learning is also significantly improved compared with traditional collaborative learning.

关 键 词：蚁群聚类算法 协作学习 分组 平均相似性 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]