基于改进蚁群算法的协作学习分组研究

协作学习中根据学习者的特征进行有效分组对于提高学习者的学习效率具有重要的作用。基于学习者的学习能力、兴趣爱好和理解水平，在基于蚁群算法的协作学习分组中，以学习者特征相似度值作为启发信息，并针对蚁群算法可能出现的早熟收敛和停滞现象，分别在初期加入判断回退机制和在中后期对启发因子及期望因子进行动态调节以保证分组结果的准确性。模拟实验结果表明该算法在分组性能及准确性上均优于传统算法。

Research of composing cooperative learning group based on enhanced ant colony optimization algorithm

The effective approach for composing cooperative learning group is very important to facilitate learners’ learning efficiency in cooperative learning environment. Based on the attributes of learning ability, interests and understanding level, this paper proposes an Enhanced Ant Colony Optimization（EACO） algorithm to compose cooperative learning group. The similarity of learners’ attributes is the stimulating information. In order to deal with the premature convergence and stagnation, a judgment-back mechanism is used at the early stage, and the stimulating factor and expectation factor is dynamically adjusted at the later period for the accuracy. The empirical results show that proposed algorithm is better than traditional algorithms in performance and accuracy.