摘要
针对提高模型的拟合效果问题,在建模的基础上,提出一种新的模型参数估计方法。结合组合预测的思想,利用BP神经网络预测算法对改进后的灰色模型做进一步的优化修正。该模型充分结合了灰色系统和人工神经网络的各自优势,弥补了单一模型预测精度低的特点,适当扩大了模型的应用范围。
In connection with the improvement of the fitting effect of modeling, this paper puts forward the improvement scheme for pa- rameters based on GM( 1,1 ) model. In this paper, according to combination prediction method , the improved prediction model based on GM( 1,1 ) is revised by the neural network, which take advantages of the superiority of GM( 1,1 ) model and neural network. This method overcome the character of low focasting accuracy of a individual forecasting model.
出处
《内蒙古农业大学学报(自然科学版)》
CAS
北大核心
2014年第3期150-153,共4页
Journal of Inner Mongolia Agricultural University(Natural Science Edition)
基金
国家自然科学基金资助项目(51175448)
河北省教育厅基金资助项目(2009159)