区间时间序列预测准确度探讨

区间时间序列在决策过程中提供重要的信息,特别是在经济发展、人口政策、规划管理或金融监管等方面,因此如何计算出预测区间的精确度成为一个重要议题。本文提出两种区间预测准确度分析的方法,通过估计预测结果的平均区间误差平方和及平均相对区间误差和,比较不同预测方法的优劣。并由预测区间与实际区间的重叠位置,充分说明预测方法所具有的有效性。这些分析预测区间准确度的方法,将为管理者提供更客观的决策空间。

On Forecasting Efficiency Evaluation for Interval Time Series

Interval time series forecasting gives important information during decision-making processes, especially on economic developments, population policies, management planning or financial controls. And how to estimate the accuracy of forecasting data is an important issue. We presented several methods of the efficiency analysis for interval forecasting. From the overlapping parts and the non-overlapping parts of the actual intervals and the forecast intervals, it should be de- fined a criterion which is more sufficient to evaluate forecasting performance for interval data. The forecast results are compared by the mean squared error of interval and mean relative interval error. Those efficiency analysis techniques can provide more objective decision space in interval forecasting to policymakers.