基于双隐含层的AGABP偏好课程预测模型  

AGA;P Preference Course Prediction Model Based on Double Hidden Layers

作　　者：洪雅敏[1] HONG Yamin(Information and Network Center,Minxi Vocational&Technical College,Longyan Fujian 364021,China)

机构地区：[1]闽西职业技术学院信息与网络中心,福建龙岩364021

出　　处：《重庆科技学院学报（自然科学版）》2022年第3期69-73,共5页Journal of Chongqing University of Science and Technology：Natural Sciences Edition

基　　金：2020年福建省中青年教师教育科研项目“基于GA-BP神经网络的线上学习需求预测模型”(JAT201286)。

摘　　要：为了预测学习者在网络教学平台中的课程偏好,采用反映学习者学习风格的4个主要指标,构建一个基于双隐含层的AGA_BP偏好课程预测模型。针对传统BP模型收敛速度慢、易陷入局部最小值的特点,应用自适应学习速率SGD优化算法对损失函数进行优化,不断增强种群适应度,并在模型中引入自适应遗传算法来调整交叉概率和变异概率,寻找最优参数。实验结果表明,AGA_BP模型的准确率、均方误差和F1分数均优于BP和GA_BP算法。In order to predict courses which are preferred by learners on the online teaching platform,an AGA_BP preference course prediction model based on double hidden layers is constructed by adopting four main indicators which reflecting the learning style of learners.Aiming at some defects of the traditional BP model such as low convergence speed,immersing local minimization,the loss function is optimized by the adaptive learning rate SGD optimization algorithm.The fitness of the population is continuously enhanced.The adaptive genetic algorithm is introduced into the model to adjust crossover probability and mutation probability for finding the optimal parameters.The experimental results show that the accuracy,mean square error and F1 score of AGA_BP model are better than BP and GA_BP algorithms.

关 键 词：双隐含层 自适应学习 自适应遗传算法 AGA_BP算法 

分 类 号：TP312[自动化与计算机技术—计算机软件与理论]