P. Ramirez Cobo, R. E. Lillo, P. De la Concepción Morales
Markovian arrival processes (MAPs) constitute a wide class of stochastic processes that allow for the modeling of non-exponential and dependent inter-arrival times. They have been considered in a wide range of real applications where the instants at which arrivals occur are recorded, but not the sequence of states of the underlying Markov chain. In this talk we aim to model a real database where not only the occurrences but also the states are partially observed. This leads us to consider a novel subclass of MAPs which shall be denoted as The clustered-states MAP. Novel theoretical properties concerning the new process as well as an application to the modeling of recurrence times in patients suffering from bladder cancer shall be illustrated.
Keywords: Markovian arrival process; recurrence times; dependence patterns
Scheduled
GT17 Stochastic Processes and their Applications III
June 7, 2022 4:50 PM
A26