基于机器学习的农产品销量区间预测

时值实施乡村振兴战略的关键时期,农村电商发展出现高峰,为加强其供应链体系建设,对农产品进行销量预测。针对影响销量的未知因素和干扰项众多的问题,提出基于K-means聚类的偏二叉树SVM销量区间预测模型。用BP神经网络时间序列对陕西大枣价格进行预测,根据质量、价格、时间三个因素,利用聚类算法对样本划分类别,偏二叉树SVM进行分类,将销量预测在聚类区间内。实验结果表明,该模型有着极高的预测精度,可用于农产品销量预测。

Forecasting sales volume range of agricultural products with machine learning

During the key period of implementing the strategy of rural revitalization,the development of rural e-commerce is peaked.In order to strengthen the construction of its supply chain system,the sales volume of agricultural products needs to be predicted.Aiming at the problem that there are many unknown factors and interferences affecting the sales volume,this paper proposes a sales volume range forecast model with K-means clustering.Using BP neural network time series to predict the price of Shaanxi jujube,according to three factors of quality,price and time,using clustering algorithm to classify the samples,and partial binary tree SVM to classify,the sales volume is predicted in the clustering interval.The experimental results show that the model has a very high prediction accuracy and can be used to forecast the sales volume of agricultural products.