Maximum spacing estimation for continuous time Markov chains and semi-Markov processes (Q2046299)

Maximum spacing (MSP) method was introduced for parameter estimation by \textit{B. Ranneby} [Scand. J. Stat. 11, 93--112 (1984; Zbl 0545.62006)] and independently by \textit{R. C. H. Cheng} and \textit{N. A. K. Amin} [J. R. Stat. Soc., Ser. B 45, 394--403 (1983; Zbl 0528.62017)]. The authors extend the method to continous time Markov chains and semi-Markov processes and consistency of the MSP estimator is proved. The authors present simulated examples to study the performance of the MSP estimator as compared to the maximum likelihood (ML) estimator and observe that the asymptotic behaviour of the MSP estimator and the ML estimator are equivalent.