Parametric and non homogeneous semi-Markov process for HIV control (Q2462662)

A Markov chain \((J_n, S_n)_{n\geq 0}\) is considered. Here \(J_n(\omega)\in E\) (finite set), and \((S_n)\) is an increasing positive sequence, tending to infinity. Let \(Z(t)= J_n\Leftrightarrow t< S_{n+1}\). This process is said to be nonhomogeneous semi-Markov process. Let \(N(t)= \sup\{n: S_n\leq t\}\). Some definitions of transition probabilities are formulated with respect to fixed \(J_{N(t)}\) and \(S_{N(t)}\). The authors derive some recurrent equations for such probabilities and write a likelihood function for them. This process is applied as a model for AIDS control. Remark: \(N(t)\) is not a Markov time with respect to filtration \(({\mathcal F}_n)\), where \({\mathcal F}_n= \sigma(J_k, S_k; k\leq n)\). Thus the results of the paper require additional assumptions.