報告題目：Bootstrapping the spectral test on adequacy check of weak VAR models
報 告 人：李木易 教授
所在單位：廈門大學
報告時間：2022年9月7日 星期三 14:00-15:00
報告地點：騰訊會議422-536-634
校内聯系人：朱複康 fzhu@jlu.edu.cn
報告摘要：We propose a Cramer-von Mises (CM) statistics for the adequacy check of weak VAR models, in which the errors are assumed to be uncorrelated but not necessarily independent. The test statistics is constructed based on the distance between the sample periodogram of residuals and a constant. In contrast to Ljung-Box-Pierce portmanteau tests in the time domain, which use the first m lags of sample correlations of residuals, the spectral tests employ correlations of infinite lags, therefore they can detect the linearility beyond m lags. We study the asymptotic null distribution of the test statistics as well as the asymptotic power against the Pitman's local alternatives. We further employ a blockwise random weighting bootstrap method to approximate critical values of the test, and justify its first-order validity. The effectiveness of the testing procedure is demonstrated via both Monte Carlo simulations and a real example analysis.
報告人簡介：李木易，2011年獲得香港大學統計學博士，現任廈門大學王亞南經濟研究院(WISE)與經濟學院統計學與數據科學系雙聘教授,博士生導師。主要研究方向為時間序列分析，長記憶過程，資産波動率建模，模型檢驗等。擔任中國概率統計學會第十一屆理事，全國工業統計學教學研究會青年統計學家協會第一屆理事，主持（含已完成）國家自然科學基金3項，教育部計量經濟學重點實驗室（廈門大學）實驗教學項目等，參與國家自科重點項目、國家社會重大項目。主講中國大學慕課《時間序列分析》。獲廈門大學2019年教學比賽翻轉課堂二等獎。為 Journal of Econometrics, Journal of Time Series Analysis, Statistics Sinica等期刊匿名審稿人。
