Using a fuzzy method to study a small sample problem of natural disaster risk assessment

摘要

A basic problem in natural disaster management is that disastrous samples are too small to be used for risk assessment using pure probabilistic methods. In this paper, the information diffusion method relevant to fuzzy information analysis is introduced for processing small samples. A reliable probability distribution can be formulated directly from incomplete fuzzy information of the small sample. An example is discussed. Results show that this method is effective for natural disaster risk assessment.