预测赋能决策:从传统模型到大数据的方案——新冠疫情趋势研判的启示

Forecasting to Empower Decision Making:A Solution from Traditional Models to Big Data’s Scheme——The Insights of Judgement of COVID-19’s Epidemic Trend

当前,公共事务治理的高度复杂性使得传统公共管理范式下的预测及辅助决策模式受到越来越大的挑战。尤其是新冠肺炎疫情的爆发对科学精准的疫情防控需求,促使对社会重大问题演化的预测问题的关注度不断加大。本文以新冠肺炎疫情这一重大突发事件为背景,以"实证同化预测与决策模型"这一典型传统预测模型为研究案例。借助对该预测模型的迭代历程、特点和应用逻辑的深描,试图从理论层面上评估传统预测模型的特点,借此进一步反思传统预测及辅助决策模式的局限性。在此基础上,结合大数据对预测及辅助决策研究在工具论、方法论以及认识论层面带来的机遇与挑战,深入剖析预测赋能决策的主旨与未来变革。从本质上审视,基于大数据驱动的预测可以为公共决策的制定提供全新的支持模式,是对传统预测的一场深刻变革,也是未来预测范式的重要转向。

The high complexity of public affairs governance has led to increasing challenges to the traditional public management paradigm of forecasting and decision making recently. In particular,the need for scientifically accurate outbreak prevention and control of the COVID-19 has promoted increased attention to the problem of predicting the evolution of socially significant issues. An inductive single case study was adopted to study a typical traditional forecasting model named Empirical Assimilation Forecasting and Decision Model.The analysis,based on a deep description of the iterative history,characteristics and application logic of this case,reveals the characteristics of the traditional forecasting model from the theoretical level,so as to further reflect on the limitations of the traditional forecasting and auxiliary decision-making model. On this basis,combined with the opportunities and challenges brought by big data on forecasting and auxiliary decision-making at the level of Instrumentalism,epistemology and methodology,this paper deeply analyzes the gist and future change of forecasting and its auxiliary decision-making. In essence,big data-driven forecasting can provide a new mode of support for public decision making,which is a profound change to traditional forecasting and an important shift in the future forecasting paradigm.