Variational inference and learning for continuous-time nonlinear state-space models
Antti Honkela, Markus Harva, Tapani Raiko and Juha Karhunen
In: PASCAL 2008 Workshop on Approximate Inference in Stochastic Processes and Dynamical Systems, May 27-29 2008, Cumberland Lodge, UK.
Inference in continuous-time stochastic dynamical models is a challenging problem. To complement existing sampling-based methods, variational methods have recently been developed for this problem. Our approach solves the variational continuous-time inference problem by discretisation that essentially reduces it to a discrete-time problem previously considered by us. While this approach is not as elegant as its competitors, our framework makes learning the model in addition to inference easy. Other extensions such as heteroscedastic models are also relatively easy to consider within this framework.