Discrete time Markov models are used in a wide variety of social sciences. However, these possess the memoryless property, which makes them less suitable for certain applications. Semi-Markov allow more flexible sojourn distributions, can accommodate duration stay effects. An overview differences and possible obstacles regarding use semi-Markov manpower planning was first given by Valliant Milk...

We investigate the statistical complexity of estimating parameters a discrete-state Markov chain kernel from single long sequence state observations. In finite case, we characterize (modulo logarithmic factors) minimax sample estimation with respect to operator infinity norm, while in countably infinite analyze problem natural entry-wise norm derived total variation. show that both cases, is go...