摘要
为提高林业产品销售价格预测的准确性,提出了基于ARIMA和LSTM混合模型的林业产品销售价格预测系统。系统设计分析和数据库设计确保了系统整体性能。通过对历史数据的预处理,提高了数据质量。利用ARIMA模型和LSTM残差预测进行价格预测,并将两者融合以得到最终预测结果。实验结果显示,该系统的预测误差控制在3以下,具有更高的准确性和较强的可靠性,可为产品价格波动的预测提供参考。
To improve the accuracy of forestry product sales price prediction,a forestry product sales price prediction system based on a hybrid model of ARIMA and LSTM is proposed.The system design analysis and database design ensure the overall performance of the system.By preprocessing historical data,the data quality has been improved.Use ARIMA model and LSTM residual prediction for price prediction,and integrate the two to obtain the final prediction result.The experimental results show that the prediction error of the system is controlled below 3,with higher accuracy and strong reliability,which can provide reference for predicting product price fluctuations.
作者
鹿瑶
陈伟
LU Yao;CHEN Wei(Nanjing University of Posts and Telecommunications,Nanjing Jiangsu 210023)
出处
《软件》
2024年第3期92-97,共6页
Software
