Author(s): Riona Michigashira; Tomoharu Hori
Linked Author(s): Tomoharu Hori
Keywords: Reinforcement learning; Flood experience; Evacuation
Abstract: In the evacuation from flood caused by heavy rain in west Japan in 2018, low percentage of evacuation by residents became severe problem. It is difficult for people who has no experience of suffering damage to decide to evacuate just in case even though government issues warning earlier. In order to make those people to evacuate when it is necessary, analysis of the effect of experience will be useful. This paper tries to simulate such effect of experience using machine learning which is modeled after a process of human beings or animals learning things through gaining experiences.
Year: 2020