基于关联面积的变权组合预测方法及其应用

变权组合预测方法是一种应用非常广泛的组合预测方法,它能够有效提高组合预测方法的预测和拟合精度.为了提高拟合精度,利用各个单项预测方法与原始数据序列在两个相邻时刻的关联面积以及关联度,给出了一种确定变权组合预测模型变权重系数的方法.最后利用该变权重的组合预测模型建立了新疆建设兵团城镇化发展水平的变权组合预测模型,结果表明这类变权组合预测方法具有较高的预测和拟合精度.

Changeable Weight Combination Forecast Model based on Area Correlation

Variable weight combined forecasting method is a widely used forecasting method,which can effectively improve the prediction and fitting accuracy of combination forecasting method.In order to improve the precision of fitting,using correlation area in two adjacent times of individual forecasting methods and the original data sequence,the variable weight coefficients are gives.Finally,the variable weight combined forecasting model of the urbanization development level of Xinjiang Construction Corps is established.The result shows that this kind of variable weight combination forecasting method has higher prediction and fitting accuracy.