Affiliations: School of Management Science and Engineering, Central University of Finance and Economics, Beijing, China
Correspondence:
[*]
Corresponding author: Jun-Hua Chen, School of Management Science and Engineering, Central University of Finance and Economics, Beijing 100081, China. E-mail: junhuachen@cufe.edu.cn.
Abstract: The deep learning algorithm is a kind of machine learning algorithm. It is based on the biological understanding of the human brain and designs a continuous iterative and abstract process in order to get the optimal data feature representation. By studying a deep nonlinear network structure, and using a simple network structure, deep Learning can achieve approximation of complex functions and show a strong ability to concentrate on the essential characteristics of the data set from a large number of non-annotated samples. Deep Belief network (DBN) is a commonly used model of deep learning, which is a Bayesian probability generation model composed of multi-layer random hidden variables. DBN can be used as a pre-training link for deep neural networks, providing initial weight for the network. An efficient learning algorithm based on this model is to train the Restricted Boltzmann Machine first, to initialize the model parameters into the better level, and then to further training and fine tuning through a small number of traditional learning algorithms such as Back Propagation (BP). This learning algorithm not only solves the problem of slow training, but also produces very good initial parameters, greatly enhances the model’s modeling capabilities. The financial market is a multivariable and nonlinear system. The DBN model can solve the problems like initial weights and so on that other prediction methods are difficult to analyze and predict. In this paper, author uses Oil Futures market price forecast as an example, to prove the feasibility of using DBN model to predict
Keywords: Deep learning, DBN algorithm, futures market
