Bernoulli race particle filters
When the weights in a particle filter are not available analytically, standard resampling methods cannot be employed. To circumvent this problem state-of-the-art algorithms replace the true weights with non-negative unbiased estimates. This algorithm is still valid but at the cost of higher variance...
| Main Authors: | , , |
|---|---|
| Format: | Conference item |
| Published: |
MLR Press
2019
|
| _version_ | 1826294427992195072 |
|---|---|
| author | Schmon, S Deligiannidis, G Doucet, A |
| author_facet | Schmon, S Deligiannidis, G Doucet, A |
| author_sort | Schmon, S |
| collection | OXFORD |
| description | When the weights in a particle filter are not available analytically, standard resampling methods cannot be employed. To circumvent this problem state-of-the-art algorithms replace the true weights with non-negative unbiased estimates. This algorithm is still valid but at the cost of higher variance of the resulting filtering estimates in comparison to a particle filter using the true weights. We propose here a novel algorithm that allows for resampling according to the true intractable weights when only an unbiased estimator of the weights is available. We demonstrate our algorithm on several examples. |
| first_indexed | 2024-03-07T03:45:28Z |
| format | Conference item |
| id | oxford-uuid:bf5055fa-1825-4005-a4ff-c8d1c6c3f088 |
| institution | University of Oxford |
| last_indexed | 2024-03-07T03:45:28Z |
| publishDate | 2019 |
| publisher | MLR Press |
| record_format | dspace |
| spelling | oxford-uuid:bf5055fa-1825-4005-a4ff-c8d1c6c3f0882022-03-27T05:46:33ZBernoulli race particle filtersConference itemhttp://purl.org/coar/resource_type/c_5794uuid:bf5055fa-1825-4005-a4ff-c8d1c6c3f088Symplectic Elements at OxfordMLR Press2019Schmon, SDeligiannidis, GDoucet, AWhen the weights in a particle filter are not available analytically, standard resampling methods cannot be employed. To circumvent this problem state-of-the-art algorithms replace the true weights with non-negative unbiased estimates. This algorithm is still valid but at the cost of higher variance of the resulting filtering estimates in comparison to a particle filter using the true weights. We propose here a novel algorithm that allows for resampling according to the true intractable weights when only an unbiased estimator of the weights is available. We demonstrate our algorithm on several examples. |
| spellingShingle | Schmon, S Deligiannidis, G Doucet, A Bernoulli race particle filters |
| title | Bernoulli race particle filters |
| title_full | Bernoulli race particle filters |
| title_fullStr | Bernoulli race particle filters |
| title_full_unstemmed | Bernoulli race particle filters |
| title_short | Bernoulli race particle filters |
| title_sort | bernoulli race particle filters |
| work_keys_str_mv | AT schmons bernoulliraceparticlefilters AT deligiannidisg bernoulliraceparticlefilters AT douceta bernoulliraceparticlefilters |