Title :
Short term forecasting of Shanghai Composite Index based on GARCH and data mining technique
Author :
Wei Shen ; Han, Yanmei
Author_Institution :
Sch. of Bus. & Adm., North China Electr. Power Univ., Beijing, China
Abstract :
Stock index forecasting is an important issue for investors and financial researchers as the movements of stock indices are nonlinear and subject to multiple factors. In this paper, we try to forecast the movements of Shanghai Composite Index using Generalized Autoregressive Condition Heteroskedasticity model. In order to increase accuracy, we introduced data mining technique and carried out forecast with single, multiple and optimized indicators. Through comparison of forecasting results, we reached the following conclusion: Forecasting results with optimized indicator groups have higher accuracy, of which the combination of MACD, PSY12 and closing indices of 2 days before has the best result.
Keywords :
autoregressive processes; data mining; forecasting theory; stock markets; GARCH; Shanghai composite index; data mining technique; generalized autoregressive condition heteroskedasticity model; optimized indicator groups; short term forecasting; stock index forecasting; Artificial neural networks; Computational modeling; Computers; Educational institutions; Indexes; Predictive models; Data Mining; GARCH; Shanghai Composite Index; Stock Index Forecasting;
Conference_Titel :
Circuits,Communications and System (PACCS), 2010 Second Pacific-Asia Conference on
Conference_Location :
Beijing
Print_ISBN :
978-1-4244-7969-6
DOI :
10.1109/PACCS.2010.5626991