大连海事大学学报 >
2020 , Vol. 46 >Issue 1: 107 - 113
DOI: https://doi.org/10.16411/j.cnki.issn1006-7736.2020.01.012
基于EEMD-PSO-LSSVM的中国沿海散货运价指数预测
收稿日期: 2019-09-05
修回日期: 2019-11-02
网络出版日期: 2019-11-02
基金资助
国家自然科学基金面上项目(71571025).
Prediction of China coastal bulk freight index based on EEMD-PSO-LSSVM
Received date: 2019-09-05
Revised date: 2019-11-02
Online published: 2019-11-02
针对中国沿海散货运价指数(CBFI)预测对精度的要求,从内在波动特性角度出发,提出一种基于集合经验模态分解(EEMD)粒子群优化算法(PSO)最小二乘法支持向量机(LSSVM)的组合预测模型.对比LSSVM、PSO-LSSVM、EMD-PSO-LSSVM三种预测模型,EEMD可对CBFI序列中波动较大数据进行降噪分解,保留序列的内在波动特性,且预测精度有一定提升,预测性能更佳.
贾红雨 , 周晨昕 , 王宇涵 , 林岩 . 基于EEMD-PSO-LSSVM的中国沿海散货运价指数预测[J]. 大连海事大学学报, 2020 , 46(1) : 107 -113 . DOI: 10.16411/j.cnki.issn1006-7736.2020.01.012
To meet the requirement of China coastal bulk freight index (CBFI) prediction accuracy, from the perspective of internal fluctuation characteristics, a combined prediction model based on ensemble empirical mode decomposition (EEMD)-particle swarm optimization (PSO)-least squares support vector machine (LSSVM) was proposed. The comparison of three prediction models of LSSVM, PSO-LSSVM and EMD-PSO-LSSVM shows that EEMD can decompose and denoise the data with large fluctuation in CBFI sequence to reserve the inherent fluctuation characteristics of the sequence. Moreover, the prediction accuracy is improved to a certain extent, and the prediction performance is better.
/
| 〈 |
|
〉 |