Bayesian Estimation of Probability of Incidences of the Most Serious Oncological Diseases in the Czech Republic

Abstract

The aim of this article is to estimate the probability of incidences of the most serious two oncological diseases in the three selected five-year age groups in the Czech Republic by using Bayesian theory of credibility, specifically binomial/beta model. These diseases are occurred in early childhood. During the lifetime, these incidences grow rapidly. The next aim of this article is to determine the parameters of a priori beta distribution, specifically in the case that the priory information on estimated parameter of binomial distribuion is known. These parameters of beta distribution are determined by using basic characteristics of beta distribution. Possibilities of permanent update of estimates can be used in commercial and health insurance companies. Data on cancer incidences are obtained from database of ÚZIS.