Coordinating Disaster Emergency Response with Heuristic Reinforcement Learning

  • 2018-11-12 21:39:07
  • Long Nguyen, Zhou Yang, Jiazhen Zhu, Jia Li, Fang Jin
  • 0

Abstract

A crucial and time-sensitive task when any disaster occurs is to rescuevictims and distribute resources to the right groups and locations. This taskis challenging in populated urban areas, due to the huge burst of help requestsgenerated in a very short period. To improve the efficiency of the emergencyresponse in the immediate aftermath of a disaster, we propose a heuristicmulti-agent reinforcement learning scheduling algorithm, named as ResQ, whichcan effectively schedule the rapid deployment of volunteers to rescue victimsin dynamic settings. The core concept is to quickly identify victims andvolunteers from social network data and then schedule rescue parties with anadaptive learning algorithm. This framework performs two key functions: 1)identify trapped victims and rescue volunteers, and 2) optimize the volunteers'rescue strategy in a complex time-sensitive environment. The proposed ResQalgorithm can speed up the training processes through a heuristic functionwhich reduces the state-action space by identifying the set of particularactions over others. Experimental results showed that the proposed heuristicmulti-agent reinforcement learning based scheduling outperforms severalstate-of-art methods, in terms of both reward rate and response times.

 

Quick Read (beta)

loading the full paper ...