摘要
高校学生的学习危机问题日趋严重,传统的管理手段和预警方法在新的学情面前显得力不从心。针对学习危机多成因和分类的特点,提出将改进的RBF神经网络用于该问题的求解。通过教师和专家对影响学习危机的因素进行分析和抽取,使用AHP层次分析法计算这些因素的权重,按权重大小重新修正主要影响因素。为获得全局最优解和提高收敛速度,利用遗传算法对传统RBF网络的权重向量进行全局搜索以得到最优模型。应用结果证明:该模型相比传统模型,在收敛速度和误差精度方面都有较大的提升,计算结果有较高的正确率和识别能力,能较好满足学习预警的实际要求。
The learning crisis of college students is becoming more and more serious.Traditional management means and early warning methods are unable to meet the new academic situation.According to the characteristics of multiple causes and classification,this paper proposes an improved RBF neural network to solve this problem.Through the analysis and extraction of factors affecting learning crisis by teachers and experts,the weights of these factors wese calculated by AHP,and the main factors were revised on the basis of the weight.In order to obtain the global optimal solution and improve the convergence speed,genetic algorithm was used to globally search the weight vectors of traditional RBF networks to obtain the optimal model.The results show that compared with the traditional model,our model has a greater improvement in convergence speed and error accuracy.And the calculation results have a higher accuracy and recognition ability,which can better meet the practical requirements of learning early warning.
作者
宋楚平
李少芹
蔡彬彬
Song Chuping;Li Shaoqin;Cai Binbin(School of Information Engineering,Nanjing Polytechnic Institute,Nanjing 210048,Jiangsu,China;College of Architectural Engineering,Jiangsu College of Engineering and Technology,Nantong 226007,Jiangsu,China)
出处
《计算机应用与软件》
北大核心
2020年第8期39-44,共6页
Computer Applications and Software
基金
教育部人文社会科学研究项目(18YJA880069)。
关键词
学习危机
指标权重
神经网络
学习预警
Learning crisis
Index weight
Neural network
Learning early warning
