基于小波神经网络的辣椒价格时间序列预测模型

摘要

By wavelet neural network intelligent analysis method, choosing pepper prices as study object, pepper retail price data from ten collection sites in Tianjin city from January 1st, 2013 to December 30th, 2013 as samples, the pepper price time series prediction mod-el to test price model was builded. Results showed that, model prediction error rate was less than 0. 01, and the correlation of predicted val-ue and actual value was 0. 8438, showing that the model could accurately predict pepper price movements. The establishment of the model would provide technical support for pepper market monitoring and early warning and references for related policies.%以辣椒价格为研究对象,用天津市10个采集点2013年1月1日—12月30日的辣椒日度市场价格数据作为样本,借助小波神经网络智能分析方法,构建了天津市辣椒价格时间序列预测模型。结果表明,模型预测误差率小于0．01,预测值与实际值之间的相关性R2为0．8438,可准确预测辣椒价格走势。该模型的构建为辣椒市场监测预警提供技术支持,也可为相关政策出台提供参考。