大数据背景下中国季度失业率的预测研究——基于网络搜索数据的分析

发布时间：2018-05-06 02:00

本文选题：失业率预测 + 大数据　；参考：《系统科学与数学》2017年02期

【摘要】：目前,中国失业率统计存在一定局限,不利于准确及时地反映劳动市场的就业变动,大数据技术的快速发展为中国失业率统计提供新的发展视角.基于网络搜索数据,文章从5种常用的预测方法中筛选出最优的支持向量机回归模型,对中国季度失业率进行了预测研究.研究表明,基于网络搜索数据预测的失业率能够比官方数据更早地反映失业趋势的变化,预测失业率与修正后的失业率水平接近,能够为政府部门提供中国失业状况的政策预警.
[Abstract]:At present, there are some limitations in the statistics of China's unemployment rate, which is not conducive to accurately and timely reflecting the employment changes in the labor market. The rapid development of big data's technology provides a new perspective for the development of China's unemployment rate statistics. Based on the network search data, this paper selects the optimal support vector machine regression model from five commonly used forecasting methods, and makes a prediction study on the quarterly unemployment rate in China. The study shows that the unemployment rate predicted on the basis of web search data is able to reflect changes in unemployment trends earlier than the official data, and that the predicted unemployment rate is close to the revised unemployment rate. To provide government departments with a policy warning of unemployment in China.
【作者单位】：东北财经大学统计学院;东北财经大学博士后科研流动站;
【基金】：国家社科基金重大项目(2015YZD08);国家社科基金项目(14CRK019) 国家自然科学基金项目(71573034) 辽宁省教育厅项目(LN2016JD020) 中国博士后科学基金(2016M601318)资助课题
【分类号】：F249.2

【相似文献】