摘要
研究区域经济预测准确率问题,区域经济变化具有高度非线性,各影响因子存在信息冗余,系统存在随机性,造成传统预测方法的预测准确率较低。为了提高区域经济预测准确率,利用数据挖掘中主成分分析的BP神经网络优势,组成一种新的区域经济预测模型。首先对区域经济影响因子进行主成份分析,消除各影响因子之间的冗余信息,减少了BP神经网络的输入变量,加快了学习速度,最后通过某地区1985-2005年经济数据对模型性能进行验证性测试,实验结果表明,新模型提高了区域经济预测的准确率,研究成果具有一定的推广和应用价值。
Regional economy changing is highly nonlinear, information redundancy exists in various influence fac- tors, and the accuracy of the traditional prediction method is lower. In order to improve the regional economic fore- casting accuracy, we used principal component analysis and BP neural network to form a new regional economic fore- casting model. The regional economic factors were used for the main component analysis, to eliminate the redundant information between each influence factors, and reduce the input variables and learning speed of BP neural network, Finally, with some region annual economic data from 1985 to 2005, the model performance was validated. The exper- imental results show that the new model improves the regional economic forecasting accuracy, and has certain promo- tion and application value.
出处
《计算机仿真》
CSCD
北大核心
2012年第6期347-350,共4页
Computer Simulation
关键词
区域经济
神经网络
数据挖掘
预测
Economy prediction
Neural networks
Data mining
Prediction
