摘要
首先分析了影响广东省第三产业发展的主要因素,指出由于上述因素相互制约、相互影响,导致第三产业的发展呈现出高度的非线性特征,并使得单一的预测模型在预测效果和泛化能力方面难以胜任.在此基础上,提出了基于神经网络集成的组合预测模型,对广东省第三产业的发展进行预测,阐述了算法的基本原理和数据处理流程,实证分析表明:基于神经网络集成的组合预测模型要比单一预测模型的预测精度高.
This paper analyzed the main factors on the improvement of tertiary industry,and pointed out that the single forecast model was difficult to satisfy the need of economic forecasting as the factors restrict and influence each other.On this basis,we proposed the combination forecasting model based on neural network ensemble,forecasted the development of tertiary industry of Guangdong Province,and described the basic principles and data processing algorithms.The empirical analysis shows that the combination forecasting model based on neural network ensemble has high prediction accuracy than a single forecast model.
出处
《经济数学》
2016年第2期50-56,共7页
Journal of Quantitative Economics
基金
通货膨胀对我国工业经济运行效率的影响机制
效应与对策研究(教育部人文社会科学研究项目
11YJC790205)
关键词
第三产业
核方法
组合预测
支持向量回归
神经网络
tertiary industry
kernel method
forecast model
support vector regression
neural network ensemble
