摘要
由于近几年大数据和机器学习的火热,也带动传统企业向数据驱动这方面转型,在餐饮业,存在着大量的历史消费数据未被科学的利用,为了能够让数据说话,通过对餐饮消费数据的分析,为了减少采购菜品的浪费和保持菜品的新鲜度,现提出一种基于LSTM模型的菜品销量预测方法。利用近几年的消费数据,结合深度学习框架(Tensor Flow)对未来的菜品销量进行预测,结果显示该模型针对时间序列数据能够很好地拟合实验数据,预测精确度符合实际需求。
Due to the fiery development of big data and machine learning in recent years, the traditional enterprises have also driven the transformation of data-driven enterprises. In the restaurant industry, a large amount of historical consumption data has not been scientifically utilized.In order to enable the data to be spoken, consumption data analysis, in order to reduce the waste of purchased dishes and to maintain the freshness of the dishes, proposes a method for forecasting sales of dishes based on the LSTM model. The consumption data in recent years are combined with the Tensor Flow to predict the future sales of dishes, the results show that the model can fit the experimental data well for the time series data and the prediction accuracy meets the actual demand.
作者
马超群
王晓峰
MA Chao-qun;WANG Xiao-feng(College of Information Engineering,Shanghai Maritime University,Shanghai 201306)
