Academic Journal of Computing & Information Science, 2023, 6(1); doi: 10.25236/AJCIS.2023.060103.
School of Information Science and Engineering, Chongqing Jiaotong University, Chongqing, 400041, China
The study of futures price forecasting is of great significance to society and enterprises. The time series characteristics of futures prices are difficult to capture because of their non-stationary and nonlinear characteristics. This paper proposes an IMAT-LSTM model, which allocates the weight of the multivariable characteristics of futures prices through an improved attention mechanism, and comprehensively compares the prediction performance with LSTM and traditional AT-LSTM models to verify the effectiveness of the model proposed in this paper in futures price prediction. The performance of the model proposed in this paper is evaluated by fitting degree indicators (MAE, RMSE, R^2). The results show that, compared with LSTM and AT-LSTM models, the fitting accuracy of the proposed model is improved by 40.31% compared with the traditional LSTM model.
Attention; LSTM; Futures price forecast; Hybrid model
Wei Wei. Research on Multivariate Futures Price Forecasting Based on IMAT-LSTM Model. Academic Journal of Computing & Information Science (2023), Vol. 6, Issue 1: 14-18. https://doi.org/10.25236/AJCIS.2023.060103.
 Faldzinski, M, Fiszeder P, Orzeszko W. Forecasting volatility of energy commodities: Comparison of GARCH models with support vector regression [J]. Energies, 2020, 14(1): 6.
 Rubio L, Alba K. Forecasting Selected Colombian Shares Using a Hybrid ARIMA-SVR Model [J]. Mathematics, 2022, 10(13): 2181.
 AHMAR A S, SINGH P K, VAN T N, et al. Prediction of BRIC stock price using ARIMA, SutteARIMA, and holt-winters[J]. Computers, Materials and Continua, 2021: 523-534.
 Santos G C, Barboza F, Veiga A C P, et al. Forecasting Brazilian Ethanol Spot Prices Using LSTM [J]. Energies, 2021, 14(23): 7987.
 HUANG N E, SHEN Z, LONG S R, et al. The empirical mode decomposition and the Hilbert spectrum for nonlinear and non-stationary time series analysis [J]. Proceedings of the Royal Society of London. Series A: mathematical, physical and engineering sciences, 1998, 454(1971): 903-995.