摘要
黄金作为一种特殊的金融商品,其价格受国际原油、美元汇率、通货膨胀等多种因素的影响,波动性强。使用单一模型进行黄金价格预测通常效果不佳,只有充分考虑价格变化的各个方面才能更加准确地预测黄金价格。应用小波分析将黄金价格分解为4个不同变化趋势,应用LS-SVM与ARIMA模型对不同变化趋势进行建模预测,并重构黄金价格组合预测的结果。实证结果表明,该组合模型预测精度比单一模型预测精度高。
Gold is a special financial product. Its price is very volatile, collaboratively influenced by the interna-tional crude oil,dollar exchange rate,inflation and other factors. One single model usually can not attain satisfactory re-sults in predicting gold price, not bringing every aspect of price changes that is essential to precise prediction into con-sideration. The paper decomposed gold price into 4 different changing trends by wavelet analysis, modeled these trends and made prediction using combined LS - SVM and A R I M A model, and reconstructed the combined gold price predic-tion results. Case study shows the combined model prediction is more accurate than single model.
作者
景志刚
施国良
Jing Zhigang Shi Guoliang(Business School of Hohai Universit)
出处
《黄金》
CAS
2017年第5期5-8,14,共5页
Gold
作者简介
景志刚（1992-）,男,江苏泰州人,硕士研究生，研究方向为数据挖掘、统计与计量;南京市西康路1号,河海大学商学院,211100;E-mail：1039713759@qq.com
