| As an important strategic reserve resource,the price fluctuation of oil has attracted much attention from the government and industry.It is of great significance to accurately predict its price fluctuation.However,the oil futures price has a strong time series effect,and the general econometric model is difficult to be applied to the out of sample prediction of oil futures price.As an important prediction method,time series prediction plays an important role in various fields.In recent years,deep learning has developed rapidly,and many scholars have developed a variety of deep learning models.Among them,neural networks that simulate neuronal thinking shine,and have made great progress in many fields,such as fitting models,image recognition,natural language processing and so on.At present,the application of deep learning model in time series prediction has become a new hotspot because of its powerful nonlinear performance.In view of the excellent performance of deep learning model in high-dimensional data,this paper studies how to use deep learning model to predict oil futures prices.The purpose of this paper is to use a variety of in-depth learning models,such as LSTM.GRU,GAN and WGAN-GP models to predict oil futures prices.This paper uses deep learning model as the general application framework of time series data prediction,then selects four specific deep learning models to predict,and analyzes and compares their results.By selecting the closing price of oil futures price as the target price,the model is trained with technical indicators,macro factors and Fourier transform as the characteristics.Finally,by comparing the prediction results of each model,it is found that the prediction results of the four deep learning models are excellent.By comparing the RMSE and Mae values of the four models,it can be found that the performance of WGAN-GP model is much better than the other three deep learning models,and the prediction accuracy of this model is the highest. |
PDF Full Text Request