Gómez Villegas, Miguel A.Marín Diazaraque, Juan Miguel2023-06-202023-06-2020020047-259X10.1006/jmva.2001.2049https://hdl.handle.net/20.500.14352/56903This paper shows a practical and easy to compute generalization of the linear dynamic model, made by assuming a continuous elliptical joint distribution for the parameters and errors. Updated distribution and probabilistic characteristics of the current and future vector of state and observations are given. As a particular simple submodel, the one with a multidimensional exponential power initial distribution is developed. An example to show its use is givenengjournal articlehttp://www.sciencedirect.com/science/journal/0047259Xopen access519.22Continuous elliptical distributionMultidimensional exponential power distributionGlobal vectorVector of statePrior distributionUpdated distributionEstadística matemática (Matemáticas)1209 Estadística