报告题目：Group LASSO for Structural Break Time Series

报 告 人：张荣茂教授

报告摘要：Consider a structural break autoregressive (SBAR) process .In practice, the number of change-points.

M is usually assumed to be known and small, because a large  m would involve a huge amount of computational burden for parameters estimation. By reformulating the problem in a variable selection context, the group least absolute shrinkage and selection operator (LASSO) is proposed to estimate an SBAR model when m is unknown. It is shown that both m and the locations of the change-points can be consistently estimated from the data, and the computation can be efficiently performed. An improved practical version that incorporates group LASSO and the stepwise regression variable selection technique are discussed. Simulation studies are conducted to assess the finite sample performance. Supplementary materials for this article are available online.

在实践中，变化点的数量m通常被认为是已知的和小的，因为一个大的m会涉及大量的参数估计计算负担。通过在变量选择上下文中重新表述问题，提出了组最小绝对收缩和选择运算符（LASSO）来估计m未知时的SBAR模型。结果表明，m和变化点的位置都可以

从数据中一致地估计，并且可以有效地执行计算。讨论了结合群LASSO和逐步回归变量选择技术的改进实用版本。进行模拟研究以评估有限样品的性能。本文的补充材料可在线获取。

报告人简介：张荣茂，浙江大学统计系教授

报告时间：  2023年5月10日14:00-16:00

报告地点:   腾讯会议:489-611-508

联 系 人：   李天然

                                                          

                                                  欢迎老师和同学参加！