A hybrid two-stage financial stock forecasting algorithm based on clustering and ensemble learning
- PDF / 4,311,432 Bytes
- 16 Pages / 595.276 x 790.866 pts Page_size
- 13 Downloads / 178 Views
A hybrid two-stage financial stock forecasting algorithm based on clustering and ensemble learning Ying Xu 1
&
Cuijuan Yang 1 & Shaoliang Peng 2 & Yusuke Nojima 3
# Springer Science+Business Media, LLC, part of Springer Nature 2020
Abstract This paper investigates the problem of the stock closing price forecasting for the stock market. Based on existing two-stage fusion models in the literature, two new prediction models based on clustering have been proposed, where k-means clustering method is adopted to cluster several common technical indicators. In addition, ensemble learning has also been applied to improve the prediction accuracy. Finally, a hybrid prediction model, which combines both the k-means clustering and ensemble learning, has been proposed. The experimental results on a number of Chinese stocks demonstrate that the hybrid prediction model obtains the best predicting accuracy of the stock price. The k-means clustering on the stock technical indicators can further enhance the prediction accuracy of the ensemble learning. Keywords Clustering . Ensemble learning . Stock price forecasting
1 Introduction Forecasting the future direction of the stock market based on the historical stock data is called the stock forecasting. The target of the stock forecasting is to propose countermeasures of the financial risk investment for corporates based on the forecast results and evaluation indicators of stock premiums. This work is based on the efficient market hypothesis [1–6], i.e. the stock market is predictable. The basic process of the financial stock forecasting is to determine the input
* Ying Xu [email protected] Cuijuan Yang [email protected] Shaoliang Peng [email protected] Yusuke Nojima [email protected] 1
College of Computer Science and Electronic Engineering, Hunan University, Changsha 410082, China
2
Computer Science, National University of Defense Technology, Changsha 410073, China
3
Graduate School of Engineering, Osaka Prefecture University, Osaka 599-8531, Japan
forecasting factors, the choice of the prediction model, and the output of the predicted results. The input technical factors include the time series based segmentation data or single data sample, and the text content characteristics related to the stock data. According to the research in the literature [7], ten representative technical indicators are selected as input factors, including Simple n-day moving average (SMA), Weighted nday moving average (EMA), Momentum (MOM), Stochastic K% (STCK), Stochastic D% (STCD), Moving average convergence divergence (MACD), Relative strength index (RSI), Larry William’s R% (WR), A/D (Accumulation/Distribution) Oscillator (ADO), Commodity Channel Index (CCI), as shown in Table 1. Prediction models include statistical models, econometric models, machine learning models, deep learning models, and hybrid models. Depending on the type of outcomes of the prediction model, the outcome variables can be divided into two categories: the amount of increase or decrease of the stock price based on t
Data Loading...
