题目: MOSUM-based test and estimation procedure for change points in panel data
摘要: In this paper, a moving sum based (MOSUM) method is proposed to detect the common breaks happened in variance of panel data. Under mild conditions, it is shown that the proposed statistic converges to an extreme distribution of a Gaussian process under the null hypothesis and diverges to infinity under the alternative hypothesis. Numerical studies are given to illustrate the performance of the proposed method. It is shown that the proposed method usually outperforms many existing cumulative sum (CUSUM) based procedures under multiple change points setting. For the single change point case, it is further shown that the proposed method performs better than the commonly used CUSUM test, for small to moderate sample size. Further, an estimation procedure for the change points is proposed based on the path of the MOSUM statistic. Application of the proposed method to the US state-level personal income data is also demonstrated.
报告人简介：王满，香港中文大学统计系博士毕业，现为东华大学管理学院讲师，主要研究方向为时间序列分析、面板数据的统计推断和空间计量分析。王满在Statistica Sinica, Economics Letter, Journal of Statistical Planning and Inference及其他统计和计量期刊上发表多篇论文，主持或参与了国家级和省部级课题多项，是Journal of Time Series Analysis, Journal of Forecasting 等国际期刊的匿名审稿人。