Summary: | 碩士 === 國立嘉義大學 === 企業管理學系 === 96 === We apply a new way called Independent Component Aalysis (ICA) from the Neurophysiological area to research the stock market volatility of 18s Taiwan weighted industrial indexes.
The Independent component analysis was introduced in Neurophysiological settings by J. Herault ,C. Jutten and B. Ans (1980), and it was usually to solve The Blind source separation problem. In this article we want to know how the effect that we combine the ICA and GARCH(1,1) model. While considering it may have structural change, we use the SWARCH model to reconstruct the market volatility .We found that in the case of combining ICA with time series model is not better than the case that not combined , at least in Taiwan industrial indexes
|