Maximizing the Computational Efficiency of Temporary Housing Decision Support Following Disasters

Postdisaster temporary housing has long been a challenging problem because of its interlinked socioeconomic, political, and financial dimensions. A significant need for automated decision support was obvious to address this problem. Previous research achieved considerable advancements in developing optimization models that can quantify and optimize the impacts of temporary housing decisions on the socioeconomic welfare of displaced families and total public expenditures on temporary housing as well as other objectives. However, the computational complexity of these models hindered its practical use and adoption by emergency planners. This article analyzes the computational efficiency of the current implementation of the most advanced socioeconomic formulation of the temporary housing problem, which uses integer programming. Moreover, it presents the development of a customized variant of the Hungarian algorithm that has a superior computational performance while maintaining the highest quality of solutions. An application example is presented to demonstrate the unique capabilities of the new algorithm in solving large-scale problems.