Parvaneh Rafieisangari
Chongqi Zhang

Liability insurance, is part of the public insurance system, which directly affects the social relationships of individuals, recognizes the rights and responsibilities of individuals towards each other and is critical to the security of businesses and professional activities. In this paper we use a non-parametric method to the estimation of distribution of liability insurance based on Maximum Entropy (ME) and then simulation of this distribution. Sampling the ME distribution is vital. Given the difficulties in carrying out exactsimulation, for the approximate simulation of the ME distribution we use Adaptive Importance Sampling (AIS) algorithm.

Liability Insurance, Maximum Entropy, Adaptive Importance Sampling