Toward Practical N2 Monte Carlo: the Marginal Particle Filter
Sequential Monte Carlo techniques are useful for state estimation in non-linear, non-Gaussian dynamic models. These methods allow us to approximate the joint posterior distribution using sequential importance sampling. In this framework, the dimension of the target distribution grows with each time...
| Hauptverfasser: | , , |
|---|---|
| Format: | Conference item |
| Veröffentlicht: |
AUAI Press
2005
|