摘要
随着原油市场环境的日益复杂,模型很难准确预测未来某段时间的原油价格。在保证预测精度的前提下为获得尽可能久的预测时长,采用模糊信息粒化方法简化计算复杂度,通过压缩样本点信息得到Up、Low和R三个模糊参数。针对原油价格时间序列周期性、非线性和长时记忆性的特点,基于支持向量机算法对模糊参数进行回归预测。研究表明,此法能够较为准确地预测未来两周的Brent原油期货价格走势和波动区间。
With the increasing complexity of the crude oil market environment,it is difficult for the model to accurately predict the crude oil price in a certain period of time in the future.In order to obtain as long prediction time as possible under the premise of ensuring the prediction accuracy,the fuzzy information granulation method is used to simplify the calculation complexity,and the up,low and R fuzzy parameters are obtained by compressing the sample information.According to the characteristics of periodicity,nonlinearity and long-term memory of crude oil price time series,the fuzzy parameters are regressed and predicted based on support vector machine algorithm.The results show that this method can accurately predict Brent crude oil futures price trend and fluctuation range in the next two weeks.
作者
张明昊
卓翔芝
ZHANG Ming-hao;ZHUO Xiang-zhi(College of Economics and Management,Huaibei Normal University,Huaibei 235000,China)
出处
《青岛大学学报(自然科学版)》
CAS
2021年第4期127-132,共6页
Journal of Qingdao University(Natural Science Edition)
基金
国家社会科学基金(批准号:15BTQ048)资助。
关键词
时间序列
模糊信息粒化
支持向量机
原油价格预测
time series
fuzzy information granulation
support vector machine
crude oil price forecasting