基于深度学习的商品销售预测研究

Research on Product Sales Prediction Based on Deep Learning

文章基于深度学习方法,通过结合粒子群优化(Particle Swarm Optimization,PSO)和长短期记忆(Long Short Term Memory,LSTM)网络,提出了一种针对大数据的商品销售预测模型。文章首先分析了LSTM的结构,其次分析了PSO方法对LSTM的优化方式,提出了PSO-LSTM商品销量预测模型,最后使用Kaggle上的数据集进行训练和测试。将所提出的模型与标准LSTM模型进行比较,结果表明,所提方法的预测精度和稳定性均优于标准LSTM方法。

Based on the deep learning method,this paper proposes a commodity sales forecasting model for big data by combining Particle swarm optimization(PSO)and Long Short Term Memory(LSTM)Network.This article first analyzes the structure of LSTM,then analyzes the optimization method of PSO method for LSTM,and proposes a PSO-LSTM product sales prediction model.Finally,this study used a dataset on Kaggle for training and testing,and compared the proposed model with the standard LSTM model.The results showed that the proposed method had better prediction accuracy and stability than the standard LSTM method.