作者ann11917 (小安)
看板NCTU-STAT100
标题[演讲] 01/04 统计所演讲公告
时间Tue Jan 1 18:57:44 2013
题 目：ESTIMATION OF MULTIVARIATE GARCH MODEL
主讲人：魏武雄教授 (Temple University, USA)
时 间：102年1月4日(星期五)上午10:40-11:30
地 点：交大综合一馆427室
Abstract
Volatility is an important issue in time series analysis. It has been heavily
investigated in many econometric and financial studies. After a brief review
of univariate volatility models, we will concentrate on multivariate general
autoregressive conditional heteroscedasticity (MGARCH) models, which have
been used to describe and forecast the time-varying variances and covariances
of economic and business data. A crucial step in the MGARCH model building is
the estimation of the linkage matrix in the model. We will propose a new
estimation method, compare it with existing procedures, discuss its merit
including its asymptotic distribution and consistency. The result will be
illustrated with both simulation and real data analysis.
--
※ 发信站: 批踢踢实业坊(ptt.cc)
◆ From: 140.113.46.59