| 基于凸组合共轭梯度法的ARIMA模型参数估计 |
| |
| 引用本文： | 李梓,单锐.基于凸组合共轭梯度法的ARIMA模型参数估计[J].数学的实践与认识,2021(1):223-229. |
| |
| 作者姓名： | 李梓 单锐 |
| |
| 作者单位： | 1.燕山大学理学院 |
| |
| 基金项目： | 风电转动系统多模态监测信息学习融合及健康预警诊断(61803329);大型工程运输车辆联合作业液压系统群协调控制理论与可靠性研究(51175448);基于时间序列分析的产品质量检测方法研究(201703A020)。 |
| |
| 摘 要： | 提出了一种凸组合共轭梯度算法,并将其算法应用到ARIMA模型参数估计中.新算法由改进的谱共轭梯度算法与共轭梯度算法作凸组合构造而成,具有下述特性:1)具备共轭性条件;2)自动满足充分下降性.证明了在标准Wolfe线搜索下新算法具备完全收敛性,最后数值实验表明通过调节凸组合参数,新算法更加快速有效,通过具体实例证实了模型的显著拟合效果.
|
| 关 键 词： | 凸组合 共轭梯度法 ARIMA模型 完全收敛 参数估计 |
Parameter Estimation Method of ARIMA Model Based on Convex Combination Conjugate Gradient |
| |
| LI Zi,SHAN Rui.Parameter Estimation Method of ARIMA Model Based on Convex Combination Conjugate Gradient[J].Mathematics in Practice and Theory,2021(1):223-229. |
| |
| Authors: | LI Zi SHAN Rui |
| |
| Institution: | (School of Science,Yanshan University,Qinhuangdao 066004,China) |
| |
| Abstract: | In this paper,a convex combination conjugate gradient algorithm is proposed and the algorithm is applied to parameter estimation of ARIMA model.The new algorithm is constructed by convex combination of improved spectral conjugate gradient algorithm and conjugate gradient algorithm,and has the following characteristics:1)The presented algorithm satisfies the conjugacy condition;2)The algorithm possesses sufficient descent property.The algorithm with standard Wolfe line search has been proved to be complete convergence.Finally,numerical experiments indicate that the new algorithm is faster and more effective by adjusting convex combination parameters,and the significant fitting effect of the model is confirmed by specific examples. |
| |
| Keywords: | convex combination conjugate gradient method ARIMA model complete convergence parameter estimation |
| 本文献已被 维普 等数据库收录！ |
