Abstract:
The article developed a model calculation of the relative frequencies of events count processes based on the use of Bayesian networks of trust. Use of the ratio of risk and the associated transition matrix is a convenient and intuitive tool for tracking changes in the processes associated with risk. Describes the need for and approaches to sampling random variables. Parameterization of the network considered for the cumulative risk of relative frequencies. Numerical examples are considered as implemented in GeNIe&SMILE which allows to design and implement graphical models probabilistic inference.
Keywords:risky behavior, last episodes, HIV-infection, risk ratio, the transition matrix, Bayesian networks.