A second-order fuzzy time series model for stock price analysis

Abstract

It is difficult to model stock market because of its uncertainty. Many methods have been introduced to tackle these difficulties, in which fuzzy time series has shown its advantages in dealing with fuzzy and uncertainty data. In recent years, many researchers have applied the fuzzy time series to analyze and forecast the stock price, and how to improve the accuracy of forecasting has attracted many researchers. In this paper, the data are first preprocessed and a new way to divide the universe of discourse is given, after which the data are fuzzified applying the triangular membership function, then three-layer back propagation (BP) neural network is established. Finally, the generalized inverse fuzzy number formula is applied to defuzzify the relation obtained with the prediction results. The proposed method is applied to predict the stock price of State Bank of India (SBI) and Dow-Jones Industrial Average (DJIA). The experimental results show that the proposed method can greatly improve the accuracy of forecasting. Furthermore, the proposed method is not sensitive to its parameters.

Keywords:

• Second-order fuzzy time series
• generalized inverse fuzzy number
• stock price
• BP neural network
• clustering analysis method

Disclosure statement

No potential conflict of interest was reported by the authors.

ORCID

Zhi Liu http://orcid.org/0000-0002-2121-3350

Additional information

Funding

This work was supported by the State Key Laboratory of Synthetical Automation for Process Industries Fundamental Research Funds, no. 2013ZCX02.