Italiano (Italian) English (Inglese)
giovedì, 14 dicembre 2017

Pubblicazioni

Indietro
Dettagli Pubblicazione
Autori:Daniele Codetta Raiteri
Luigi Portinale
Area Scientifica:Artificial Intelligence
Uncertain Reasoning
Probabilistic Graphical Models
Formal Models
Titolo:Generalized Continuous Time Bayesian Networks and their GSPN Semantics
Apparso su:Proceedings of the European Workshop on Probabilistic Graphical Models
Pagine:105-112
Editore:HIIT
Anno:2010
Tipo Pubblicazione:Paper on Proceedings International Conference
URL:http://www.helsinki.fi/pgm2010/papers/codetta.pdf
Sommario:We present an extension to Continuous Time Bayesian Networks (CTBN) called Generalized CTBN (GCTBN). The formalism allows one to model continuous time delayed variables (with exponentially distributed transition rates), as well as non delayed or “immediate” variables, which act as standard chance nodes in a Bayesian Network. The usefulness of this kind of model is discussed through an example concerning the reliability of a simple component-based system. The interpretation of GCTBN is proposed in terms of Generalized Stochastic Petri Nets (GSPN); the purpose is twofold: to provide a well-defined semantics for GCTBNin terms of the underlying stochastic process, and to provide an actual mean to perform inference (both prediction and smoothing) on GCTBN.