论文部分内容阅读
在分析学习率自适应算法和附加动量项算法优缺点的基础上,采用一种融合自适应学习率和附加动量的改进算法应用于大学生学习成绩评价模型中。通过Matlab平台仿真,结果表明基于改进算法的评价模型能够完成对大学生学习成绩的评价,且在收敛速度和评价精度方面优于标准BP算法、学习率自适应算法和附加动量项算法。
Based on the analysis of the advantages and disadvantages of learning rate adaptive algorithm and additive momentum algorithm, an improved algorithm based on adaptive learning rate and additional momentum is applied to the evaluation model of college students’ academic performance. Through the simulation of Matlab platform, the result shows that the evaluation model based on the improved algorithm can evaluate the academic performance of college students better than the standard BP algorithm, learning rate adaptive algorithm and additional momentum term algorithm in terms of convergence rate and evaluation accuracy.