基于灰色关联度的小样本预测模型

Small Sample Prediction Model based on Grey Incidence Degree

针对传统灰色预测模型无法处理以及难以寻找统计规律的小样本数据,通过对序列之间灰色关联度物理含义的分析,挖掘序列内部以及序列之间的数据变化规律,并在此基础上构建了一种新的灰色预测模型;应用该模型对中国2008年度的GDP进行预测,预测结果验证了该模型的有效性及实用性;该模型的提出为小样本数据提供了一种新的预测方法。

For a small sample data, which is not disposesd by a traditional grey prediction model and has not a statistical laws. First, by mining and analysing grey incidence degree＇s physical meaning among sequences, it can discovere change laws among or between sequences, and then a novel grey prediction model is set up. Ultimately, The GDP in 2008 is forecasted with this model, the results prove the validity and practicability. For small sample, this model puts forward a new prediction approach.