摘要
介绍一种利用遗传程序设计的方法来自动生成统计预测模型,并进行误差估计分析,改变过去只使用拟合曲线粗糙、预测结果不理想的几种传统固定统计预测模型的传统分析方法。通过对[1]和[2]的真实历史资料验证,结果表明,与传统的线性回归、指数回归、抛物线回归[3]三种方法对比,遗传程序设计建立的模型所预测的数据准确度明显要高。
An application of genetic programming in statistical modeling is proposed, which obviously improved the traditional regular methods of statistical modeling that could only obtain rough curve fitting and unsatisfactory results. Then, the estimating standard error and forecasting standard error were calculated and analyzed. By using the actual historical data from Statistics Yearbook of China and Statistics Yearbook of Jiangxi Province, China published in recent years, the automatic generated statistical model of economic forecasting by using genetic programming was established and the result indicates that the accuracy calculated by this statistical model is obviously much higher, compared with traditional methods such as linear regression, exponential regression and parabolic regression[3].
出处
《系统仿真学报》
EI
CAS
CSCD
北大核心
2005年第7期1597-1600,共4页
Journal of System Simulation
基金
国家自然科学重点基金项目(60133010)
高等学校博士学科点专项科研基金项目(20030486049)
关键词
遗传程序设计
自动建模
回归
统计预测
evolutionary computation
automatic building model
regression
statistical forecasting