Financial Trend Prediction Based on Deep Belief Network

Abstract

In order to further strengthen the control of financial market trends, a financial trend prediction model based on deep belief network (DBN) is proposed to further improve the prediction level of financial trend. Among them, the prediction and classification of financial market trend is realized by introducing Elliott wave theory. The prediction model adopts deep belief network model. Experimental results show that by introducing the Elliott wave theory, the designed financial trend prediction model based on deep belief network can achieve the accurate prediction of financial trend, the prediction precision is 67.5%, and the corresponding mean square error is 0.413. Compared with BP network and MLP network, deep belief network shows better performance on four evaluation indicators, namely ER, MAE, RMSE and MSE, and is more suitable for the design of financial trend prediction model. The above experimental results verify the feasibility and superiority of the financial trend prediction model based on deep belief network proposed in this study, which has certain application value.