改进的BP神经网络预测高校毕业生就业能力  

Prediction of College Graduates' Employ-Ability Based on the Improved BP Neural Network

作　　者：董秀英[1] 柏小丽[2] 汤铁群[1] 

机构地区：[1]阜阳师范学院物理与电子工程学院,安徽阜阳236037 [2]四川省泸州电业局,四川泸州646000

出　　处：《宜宾学院学报》2015年第6期93-96,共4页Journal of Yibin University

基　　金：阜阳师范学院教学研究项目(2012JYXM91);阜阳师范学院硕士科研启动资金

摘　　要：针对传统BP算法的神经网络收敛较慢的缺点,提出基于L-M(Levenberg-Marquard)算法的神经网络对影响高校毕业生就业能力的各项因素进行分析的模型,以预测高校毕业生的就业能力存在的不足.因素量化采用模糊数学中的格贴近度和专家打分的方法.仿真结果表明:系统模型缩短了训练时间,具有较高的准确性,预测值与实际值的误差很小,可以将此模型应用于对高校毕业生就业能力的预测.An improved neural network model based on L-M（Levenberg-Marquard） algorithm neural network was ap-plied to analyze the factors influencing college graduates＇ employability. It can remedy the shortcoming of the slow con-vergence rate of traditional BP algorithm neural network. This model can predict the deficiencies in employability of col-lege graduates. The lattice fuzzy close-degree assessment and expert scoring method were adopted, which quantified thevarious data of the factors on the employability of college graduates. Simulation results indicate that this model can re-markably reduce the training time, with relatively high accuracy. With small errors in the predicted value and actual val-ue, this network model can be used to forecast the employability of college graduates.

关 键 词：L-M算法 BP神经网络 格贴近度 就业能力 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]