Rao-blackwellised particle filtering via data augmentation
In this paper, we extend the Rao-Blackwellised particle filtering method t o more complex hybrid models consisting of Gaussian latent variables and discrete observations. This is accomplished by augmenting the models with artificial variables that enable us to apply Rao-Blackwellisation. Other impro...
| প্রধান লেখক: | , , |
|---|---|
| বিন্যাস: | Conference item |
| প্রকাশিত: |
Neural information processing systems foundation
2002
|
| সংক্ষিপ্ত: | In this paper, we extend the Rao-Blackwellised particle filtering method t o more complex hybrid models consisting of Gaussian latent variables and discrete observations. This is accomplished by augmenting the models with artificial variables that enable us to apply Rao-Blackwellisation. Other improvements include the design of an optimal importance proposal distribution and being able to swap the sampling an selection steps to handle out liers. We focus on sequent ial binary classifiers t hat consist of linear- combinations of basis functions, whose coefficients evolve according t o a Gaussian smoothness prior. Our results show significant improvements. |
|---|