The Case for Continuous Time

682
22.7
Опубликовано 28 июля 2016, 22:48
Time is a continuous quantity. This talk begins with theoretical and experimental problems that arise when time is treated as a discrete quantity in stochastic systems. I will then discuss continuous time Bayesian networks (CTBNs), a variable-based representation of continuous-time Markov processes. I will cover their representation and semantics and a bit about inference and learning in the models. Finally, I will present my group's recent work in employing CTBNs on real-world applications.
автотехномузыкадетское