【摘要】 针对协作学习异质群组构建的原则和多目标优化等问题,提出一种改进的密度聚类算法,并在此基础上进一步实现群组构建算法。求得学习者的距离矩阵,根据聚类算法对学习者进行聚类,通过基本聚类中学习者随机选择分组,用调节聚类对分组进行调节,实现异质分组。实验结果表明,该算法能够满足"组内异质,组间同质"的要求,在分组效果上优于随机选择法,在执行效率上优于遗传算法。更多还原

【Abstract】 For certain principles of heterogeneous group formation in collaborative learning and to solve multi-objective optimization problems,an improved density clustering algorithm was proposed,and based on this,agroup formation algorithm was realized.The distance matrix of learners was got and the learners were clustered.After building the basic groups by random selecting learners from each basic cluster,the groups were adjusted using adjustment clusters to realize heterogeneous group formation.Experimental results show that the algorithm can meet the requirement of group homogeneity between the heterogeneous groups,and it is better than random algorithm on effectiveness and is more efficient than genetic method algorithm.更多还原