摘要
针对服装销售过程中供求不平衡等问题,提出了一种自回归移动平均(ARIMA)和长短期神经网络(LSTM)相结合的服装流行趋势预测模型。首先,利用网络爬虫技术获取服装的销售、评论等相关信息;然后,采用ARIMA模型对不同色彩属性下的服装销量预测其线性特征,通过LSTM修正ARIMA模型的预测残差;最后,对比ARIMA模型和ARIMA-LSTM组合模型的预测精度和性能。实验表明,与单一的时间序列算法相比,结合两种高效的模型算法在服装销量预测精度上取得了较好的效果。
Aiming at the imbalance of supply and demand in the clothing sales process,a clothing trend prediction model combining autoregressive moving average(ARIMA)and long-term and short-term neural network(LSTM)is proposed in this paper.First of all,use Web crawler technology to obtain clothing sales,reviews and other related information.Then,the ARIMA model is used to predict the linear characteristics of clothing sales under different color attributes,and the prediction residuals of the ARIMA model are corrected by LSTM.Finally,compare the prediction accuracy and performance of the ARIMA model and the ARIMA-LSTM combined model.Experiments show that,compared with a single time series algorithm,the combination of two efficient model algorithms has achieved better results in the accuracy of clothing sales prediction.
出处
《工业控制计算机》
2022年第8期115-117,共3页
Industrial Control Computer
基金
国家重点研发计划资助项目(2018YFB1700702)
安徽农业大学校级自然科学基金资助项目(k2048004)。
