Deep learning with stock indicators and two-dimensional principal component analysis for closing price prediction system

Abstract

The stock market is an important component in the current economic market. And stock price prediction has recently garnered significant interest among investment brokers, individual investors and researchers. In general, stock market is very complex nonlinear dynamic system. Accordingly, accurate prediction of stock market is a very challenging task, owing to the inherent noisy environment and high volatility related to outside factors. In this paper, we focus on deep learning method to achieve high precision in stock market forecast. And a deep belief networks(DBNs), which is a kind of deep learning algorithm model, coupled with stock technical indicators(STIs) and two-dimensional principal component analysis((2D) <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> PCA) is introduced as a novel approach to predict the closing price of stock market. A comparison experiment is also performed to evaluate this model.