Data-driven decision-making model for determining the number of volunteers required in typhoon disasters

Abstract

Volunteer teams provide valuable support after large-scale disasters. However, excessive volunteer participation poses challenges for formal operations. Therefore, an appropriate decision-making method is required to quickly determine the number of volunteers required after a disaster. This study proposes a data-driven decision-making (D3M) method for typhoon disaster volunteerism that can effectively predict the number of volunteers required. Disaster data from actual cases were gathered, analyzed, and preprocessed to prepare the model. Feature selection, D3M model training and optimization, and model validation were performed to fine-tune the volunteer participant predictions. Using data from an actual typhoon in the Philippines, the rationality and efficacy of the method were verified through a comparative analysis of the experimental results. The proposed method learns from disaster-event data to quickly predict the number of volunteers needed, such that it not only reasonably allocates volunteers to assist professional teams in rescue but also avoids secondary problems caused by an overwhelming response.

Keywords

• Data-driven decision-making
• Optimization
• Rescue
• Typhoon