| Speaker: | Prof. Changliang Zou, Nankai University |
| Host: | 陈钊, School of Data Science, Fudan University |
| Time: | 15:45-16:30, Oct 24, 2019 |
| Location: | Zibin N102, Fudan University |
| Abstract: | In multiple change-point analysis, one of the main difficulties is to determine the number of change-points. In this talk, I will introduce a data-driven approach based on an order-preserved sample-splitting strategy. Under a unified framework, a cross-validation estimation scheme is developed to achieve consistent selection. Furthermore, we construct a simple yet effective selection procedure which is able to quantify “uncertainty”, say controlling the false discovery rate. The proposed methods are applicable to most kinds of popular change-point algorithms. |
| Bio: | 邹长亮 博士 ，南开大学统计与数据科学学院教授、博导。08年于南开大学获博士学位，随后留校任教。研究方向是统计学、质量科学及其与数据科学领域的交叉研究和实际应用。研究兴趣包括：高维数据统计推断、大规模数据分析、统计过程控制、变点和异常点检测等，在统计学和质量科学领域杂志上发表论文几十篇。 |