基于数据驱动的高校毕业生就业率预测研究  

Research on data⁃driven based prediction of college graduate employment rate

作　　者：刘小杰[1] LIU Xiaojie(Suqian University,Suqian 223800,China)

机构地区：[1]宿迁学院,江苏宿迁223800

出　　处：《现代电子技术》2021年第5期127-131,共5页Modern Electronics Technique

基　　金：江苏省教育科学“十三五”规划2020年课题(X—a/2020/14)。

摘　　要：高校毕业生就业率的影响因素很多,是复杂多变的,传统预测方法无法描述高校毕业生就业率的变化趋势,不仅使得高校毕业生就业率预测精度低,而且预测消耗的时间长。为了提升高校毕业生就业率预测效果,提出数据驱动的高校毕业生就业率预测方法。首先,对当前高校毕业生就业率常用预测方法进行分析,找到这些方法的弊端;然后收集高校毕业生就业率历史数据,采用支持向量机对其进行预测,找到高校毕业生就业率变化趋势,并引入蚁群算法对支持向量机的参数进行优化;最后,采用Matlab 2019编程实现高校毕业生就业率预测实验。结果表明,无论是高校毕业生就业率的预测精度或者是预测消耗时间,数据驱动方法均明显优于当前经典高校毕业生就业率预测方法,具有更加广泛的应用前景。The college graduate employment rate is influenced by multiple factors,which is complex and changing,so the traditional prediction methods fails to describe the changing trend of the employment rate,which makes the prediction accuracy of the employment rate low and the prediction duration long.In view of the above,a data⁃driven prediction method of college graduate employment rate is proposed to improve the prediction effect.The current prediction methods of college graduate employment rate is analyzed to find out their disadvantages first,and then the historical data about the employment rate are collected.Support vector machine is used to predict the rate and find out its variation trend.And then,ant colony algorithm is introduced to optimize the parameters of the support vector machine.Experiment of college graduate employment rate prediction was performed on Matlab 2018.The experimental results show that,in comparison with the current classic prediction methods of college graduate employment rate,the data⁃driven method is significantly higher in prediction accuracy and shorter in time consumption of prediction.Therefore,it has a broad application prospect.

关 键 词：就业率预测 变化趋势 预测模型 支持向量机 蚁群算法 消耗时间 

分 类 号：TN911.1-34[电子电信—通信与信息系统] TP393[电子电信—信息与通信工程]