Position: Home > Articles > A novel forestry investment forecast method based on improved support vector machine
Journal of Agricultural University of Hebei
2012,35
(4)
123-126
基于改进支持向量机的林业资金投资预测方法
作 者：
陶佳;沈红岩;高冠东
单 位：
中央司法警官学院信息管理系;河北农业大学信息科学与技术学院
关键词：
林业资金投资;回归预测;时间序列;支持向量机;粒子群算法
摘 要：
针对林业资金投资变化的定量预测,提出一种基于改进支持向量机的预测方法。利用滑动时间窗口方法将历年林业资金投资数据构造成时间序列,将其做为数据样本集并由改进支持向量机加以训练以得到预测模型。通过某省近20年的林业资金投资数据实验验证了预测方法的有效性,实验结果表明:与传统预测方法相比,基于改进支持向量机的预测方法明显提高了投资变化预测精度。
译 名：
A novel forestry investment forecast method based on improved support vector machine
作 者：
TAO Jia1,SHEN Hong-yan1,GAO Guan-dong2(1.College of Information Science and Technology,Agricultural University of Hebei,Baoding 071000,China; 2.Department of Information and Management,Central Institute for Correctional Police,Baoding 071000,China)
关键词：
forestry investment forecast;regression forecast;time series;support vector machine;particle swarm optimization
摘 要：
Focused on forecast investment forecast,a novel method based on improved support vector machine with particle swarm optimization(SVM-PSO) was proposed,in order to improve forecast accuracy.With sliding time window,history of forestry investment data had been constructed into a time series.In order to obtain forecast model,the time series was trained as sample set by SVM-PSO.Finally,the experiments,whose data is from the forestry investment of nearly 20 years,show that the SVM-PSO forecast method has better performance than the traditions.
导出引用
计量
文章访问数: 5
HTML全文浏览量: 0
PDF下载量: 0
所属期刊
