摘要
非参数模型具有不对数据分布作任何假设、不受数据类型和尺度限制、灵活度较高等特点,因此在经济学研究中被广泛应用。文章在对非参数模型基本原理进行介绍的基础上,梳理了常用的非参数模型的研究方法与工具,并归纳了不同方法的应用特点及选择依据;从宏观经济数据预测、金融市场监测、临近预测、因果推断、市场环境分析、时空数据分析、其他领域7个方面,基于近年来国内外高引用文献对非参数模型在经济学中的应用进行了介绍。
Non-parametric models have the characteristics of not making any assumptions about data distribution,not being restricted by data types and scales,and having relatively high flexibility.Therefore,they are widely used in economic research.This paper is based on the introduction of the basic principles of non-parametric models to sorts out the commonly used research methods and tools of non-parametric models,and summarize the application characteristics and selection basis of different methods.And then,based on highly cited domestic and foreign literature in recent years,the paper introduces the application of non-parametric models in economics from seven aspects:macroeconomic data monitoring,financial market transactions and monitoring,proximity prediction,causal inference,market environment analysis,spatiotemporal data analysis,and other fields.
作者
刘莹
Liu Ying(Institute of Strategy Research for the Guangdong-Hong Kong-Macao Greater Bay Area;Guangdong Academy of Sciences,Guangzhou 510070,China)
出处
《统计与决策》
北大核心
2025年第14期35-40,共6页
Statistics & Decision
基金
广东省科学院打造综合产业技术创新中心行动资金项目(2022GDASZH-2022030601)。
关键词
参数
非参数模型
机器学习
风险研判
因果推断
parameter
non-parametric model
machine learning
risk assessment
causal inference
