基于LSTM的棉花期货价格预测研究

Research on Prediction of Cotton Futures Price Based on LSTM

摘要: 本文主要基于LSTM神经网络对棉花期货价格进行预测研究。结果表明，LSTM模型预测的均方根误差(RMSE)、平均绝对误差(MAE)以及平均绝对百分比误差(MAPE)均小于Xgboost模型，故LSTM的预测精度比XGboost更好。LSTM模型在棉花期货价格上的预测表现出较好的性能。

Abstract: In this paper, LSTM neural network is used to predict cotton futures prices. The experiment results show that the root mean square error (RMSE), mean absolute error (MAE) and mean absolute per-centage error (MAPE) of the LSTM model were smaller than the Xgboost model, so the forecast ac-curacy of LSTM model is better than Xgboost model, and LSTM model has good performance in the prediction of cotton futures prices.