Hits:

Indexed by:期刊论文

Date of Publication:2013-01-01

Journal:International Journal of Information Processing and Management

Included Journals:Scopus

Volume:4

Issue:3

Page Number:215-221

ISSN No.:20934009

Abstract:With the development of economy and the improvement of people's life, more and more people are involved in the stock market. Therefore, stock price forecasting is very important that it has practical significance in both the financial supervision of the government and the prevention of investors' risk on stock market. However, stock price is affected by a lot of factors, so stock price series is complex, nonlinear and dynamic that it's difficult to predict it effectively by a single method. The ARMA (autoregressive and moving average) model is one of the most popular and widely-used time series model that can predict linear problem, while BPNN (back propagation neural network) is commonly used to process nonlinear problem. This paper proposes a combined model of ARMA, BPNN and Markov model to forecast the stock price. ARMA and BPNN are used to solve the linear and nonlinear component of the stock price series respectively and Markov model can modify the result and make it more accurate. The experimental results show that ARMA-BPNN model outperforms single ARMA or BPNN model, while ARMA-BPNN-Markov model gets more accurate result than ARMA-BPNN model.