A Hierarchical Beta Process Approach for Financial Time Series Trend Prediction
An automatic stock market categorization system would be invaluable to investors and financial experts, providing them with the opportunity to predict a stock price changes with respect to the other stocks. In recent years, clustering all companies in the
- PDF / 2,179,240 Bytes
- 11 Pages / 439.37 x 666.142 pts Page_size
- 2 Downloads / 171 Views
2
School of Computer Science and Engineering, University of New South Wales, Sydney, Australia {mojgang,wong}@cse.unsw.edu.au National ICT Australia (NICTA), Sydney, Australia {fang.chen,yang.wang}@nicta.com.au 3 Novel Approach Limited, Hong Kong, China [email protected]
Abstract. An automatic stock market categorization system would be invaluable to investors and financial experts, providing them with the opportunity to predict a stock price changes with respect to the other stocks. In recent years, clustering all companies in the stock markets based on their similarities in shape of the stock market has increasingly become popular. However, existing approaches may not be practical because the stock price data are high-dimensional data and the changes in the stock price usually occur with shift, which makes the categorization more complex. In this paper, a hierarchical beta process (HBP) based approach is proposed for stock market trend prediction. Preliminary results show that the approach is promising and outperforms other popular approaches. Keywords: Stock trend prediction GARCH-based clustering
1
·
Hierarchical beta process
·
Introduction
Time series analysis is finding the correlations in data. The most significant challenges faced by time series analysts include defining the shape of time series, forecasting their future trends and classifying them to different categories. Solving these tasks can significantly help the economy and society. These tasks are actively applied in the stock market section. Malkiel and Fama believe that historical stock prices can be efficiently used to predict their future trend [5]. However, stock market analysis is a complicated task due to the large, high-dimensional, non-normally-distributed and non-stationary time series. Nonstationary means that the statistical properties such as mean, variance and/or autocovariance, change over time. New theories and methods are needed to handle these settings. In particular, forecasting a stock market price movement trend has been a great challenge for a long time. Lots of works have been done on mining the financial time series [12,13,15,16]. c Springer International Publishing Switzerland 2016 H. Cao et al. (Eds.): PAKDD 2016 Workshops, LNAI 9794, pp. 227–237, 2016. DOI: 10.1007/978-3-319-42996-0 19
228
M. Ghanavati et al.
Huang et al. [6] used a wrapper approach to choose the best feature subset of 23 technical indices and then combined different classification algorithms to predict the future trend of Korea and Taiwan stock markets. Yu et al. [19] proposed a support vector machine (SVM) based model called least squares support vector machine (LSSVM) to predict stock market trends. They have used genetic algorithm, first to select the input features for LSSVM and then to optimize LSSVM parameters. To evaluate efficiency of the proposed model, S&P 500 index, Dow Jones industrial average (DJIA) index, and New York stock exchange (NYSE) index, have been used as testing targets. Patel et al. [14] compared the application of four models
Data Loading...
