Quantifying post-disaster business recovery through Bayesian methods
Quantifying post-disaster business recovery through Bayesian methods
复制标题
DOI：
10.1080/15732479.2020.1777569
复制
发表时间：
2020-06-12
影响因子：
3.7
通讯作者：
Xiao, Yu
中科院分区：
文献类型：
--
作者：
Aghababaei, Mohammad;Koliou, Maria;Xiao, Yu
灾后企业恢复在社区的社会经济恢复中发挥着重要作用。本研究的重点是发展的概率建模方法，通过贝叶斯线性回归量化和预测业务恢复。建议的建模方法包括三个步骤，包括数据收集，模型形式的发展，并通过严格的评估和消除步骤的模型选择。考虑了描述企业灾后恢复状态的四个属性，即业务停止运营天数、收入恢复、客户保留和员工保留。这项研究的主要贡献之一是将家庭和企业之间的相互作用，在社区开发预测业务恢复模型。朝着这个方向，不同的方法来考虑到家庭恢复到客户留存率的业务的影响进行了调查和建议。作为应用，将所提出的建模方法应用于北卡罗来纳州兰伯顿社区的纵向实地研究结果，该社区受到2016年飓风马修的严重影响，重点关注业务恢复。本研究所提出的预测模型可进一步应用于灾害事件后基于风险的社区恢复力评估。
Business recovery after a disaster plays an important role in the socioeconomic recovery of a community. This study focuses on the development of a probabilistic modelling approach for quantifying and predicting business recovery through Bayesian linear regression. The proposed modelling approach consists of three steps including data collection, development of model forms, and model selection through rigorous evaluation and elimination steps. Four attributes, namely business cease operation days, revenue recovery, customer retention, and employee retention, which describe the post-disaster recovery state of a business, are considered. One of the main contributions of this study is incorporating the interplay between household and businesses in a community in developing predictive business recovery models. Towards that direction, different methods to account for the effect of household recovery into the customer retention rate of a business are investigated and proposed. As an application, the proposed modelling approach is applied on the results of a longitudinal field study at the community of Lumberton, NC, which was heavily impacted by the 2016 Hurricane Matthew, focusing on business recovery. The predictive models proposed in this study may be further applicable in risk-based resilience assessment of communities following disastrous events.