报告承办单位:数学与统计学院

报告题目:Burn-in selection in simulating time series

报告内容:Many time series models are defined in a recursive manner, which prohibitsexact simulations. In practice, one appeals to simulating a long time series anddiscarding a large number of initial simulated observations, known as the burn-in.For autoregressive models where the dependence decays exponentially fast, the choice ofthe burn-in is not critical. However, for long-memory time series where the dependence
from the remote past is strong, it is not clear how to select the burn-in number. Bycombining several samplers with randomized burn-in numbers, we develop a method forexactly simulating the expectation of a statistic computed from a time series. Moreover,with some suitably chosenstatistics, the exact simulation method can be applied toquantify the effect of burn-in numbers on the simulated sample. Extensive simulationstudies are conducted to provide some practical guidance for burn-in selections.

报告人姓名:Chun Yip Yau

报告人所在单位:ChineseUniversityof Hong Kong

报告人职称:教授

报告时间:2022年12月5日, 星期一，下午3:00-4:00

在线报告:腾讯会议号码：236-568-028

报告人简介：Chun Yip Yau is currently a professor in Department of Statistics, ChineseUniversity of Hong Kong. He obtained his PhD from Columbia University in 2010. Hisresearch interest includes time series, change-point analysis, spatial statistics andextreme value theory.