摘要
Lasso类方法可以同时实现变量选择与参数估计,将之运用于AR(p)模型的定阶及参数估计,可以大大简化计算步骤和时间.本文在前人基础上利用Lasso类方法,改进了AR(p)模型的定阶与参数估计,通过计算机编程模拟,验证了此类方法的可行性,并比较了在不同样本量情况下,Lasso和Adaptive Lasso方法在定阶和参数估计两方面的优良性,最后将较优的Adaptive Lasso方法用于实际时间序列数据中,并对结果进行分析,指出了该方法的实用性.
Lasso method can realize the variable selection and parameter estimation at the same time.This method can be applied to the problem of order determination and parameter estimation for AR（p）model,which can greatly simplify the calculation steps and times.Based on the studies of predecessors,this study can improve the parameter estimation and order determination for AR（p）model by means of Lasso method.Through computer simulation the feasibility of this method is verified.This article also compare the performance of Lasso and Adaptive Lasso method in different samples,then choose the Adaptive Lasso method for real time series data.As results are analyzed,the practicability of this method is pointed out.
出处
《浙江工业大学学报》
CAS
2014年第4期463-467,共5页
Journal of Zhejiang University of Technology
基金
国家自然科学天元基金资助项目(11326126)
作者简介
谢仪（1990-），女，新疆乌鲁木齐人，硕士研究生，研究方向为数理统计，E-mail：822122040@qq．com．
通信作者：景英川副教授，E—mail：184689432@qq．com．
