论文标题
在输入输出网络中重建需求冲击
Reconstruction of Demand Shocks in Input-Output Networks
论文作者
论文摘要
投入输出分析描述了部门和地区之间的生产，需求和贸易的依赖性，并允许通过经济网络了解经济冲击的传播。实际应用中的一个核心挑战是数据的可用性。观察结果可能仅限于少数部门的冲击的影响，但是对起源和影响的完整图片非常需要指导政治对策。在本文中，我们证明，可以完全从对生产变化的有限观察结果中完全重建最终需求的最终需求。我们从稀疏信号恢复中调整了三种算法，并评估了它们的性能和鲁棒性到观察不确定性。
Input-Output analysis describes the dependence of production, demand and trade between sectors and regions and allows to understand the propagation of economic shocks through economic networks. A central challenge in practical applications is the availability of data. Observations may be limited to the impact of the shocks in few sectors, but a complete picture of the origin and impacts would be highly desirable to guide political countermeasures. In this article we demonstrate that a shock in the final demand in few sectors can be fully reconstructed from limited observations of production changes. We adapt three algorithms from sparse signal recovery and evaluate their performance and their robustness to observation uncertainties.
