The infinite factorial hidden Markov model
We introduce a new probability distribution over a potentially infinite number of binary Markov chains which we call the Markov Indian buffet process. This process extends the IBP to allow temporal dependencies in the hidden variables. We use this stochastic process to build a nonparametric extensio...
| Main Authors: | , , |
|---|---|
| Format: | Journal article |
| Language: | English |
| Published: |
2009
|
| _version_ | 1797102981357764608 |
|---|---|
| author | Van Gael, J Teh, Y Ghahramani, Z |
| author_facet | Van Gael, J Teh, Y Ghahramani, Z |
| author_sort | Van Gael, J |
| collection | OXFORD |
| description | We introduce a new probability distribution over a potentially infinite number of binary Markov chains which we call the Markov Indian buffet process. This process extends the IBP to allow temporal dependencies in the hidden variables. We use this stochastic process to build a nonparametric extension of the factorial hidden Markov model. After constructing an inference scheme which combines slice sampling and dynamic programming we demonstrate how the infinite factorial hidden Markov model can be used for blind source separation. |
| first_indexed | 2024-03-07T06:13:31Z |
| format | Journal article |
| id | oxford-uuid:f056a31a-2825-4e30-8b2b-1c08d95f0391 |
| institution | University of Oxford |
| language | English |
| last_indexed | 2024-03-07T06:13:31Z |
| publishDate | 2009 |
| record_format | dspace |
| spelling | oxford-uuid:f056a31a-2825-4e30-8b2b-1c08d95f03912022-03-27T11:47:00ZThe infinite factorial hidden Markov modelJournal articlehttp://purl.org/coar/resource_type/c_dcae04bcuuid:f056a31a-2825-4e30-8b2b-1c08d95f0391EnglishSymplectic Elements at Oxford2009Van Gael, JTeh, YGhahramani, ZWe introduce a new probability distribution over a potentially infinite number of binary Markov chains which we call the Markov Indian buffet process. This process extends the IBP to allow temporal dependencies in the hidden variables. We use this stochastic process to build a nonparametric extension of the factorial hidden Markov model. After constructing an inference scheme which combines slice sampling and dynamic programming we demonstrate how the infinite factorial hidden Markov model can be used for blind source separation. |
| spellingShingle | Van Gael, J Teh, Y Ghahramani, Z The infinite factorial hidden Markov model |
| title | The infinite factorial hidden Markov model |
| title_full | The infinite factorial hidden Markov model |
| title_fullStr | The infinite factorial hidden Markov model |
| title_full_unstemmed | The infinite factorial hidden Markov model |
| title_short | The infinite factorial hidden Markov model |
| title_sort | infinite factorial hidden markov model |
| work_keys_str_mv | AT vangaelj theinfinitefactorialhiddenmarkovmodel AT tehy theinfinitefactorialhiddenmarkovmodel AT ghahramaniz theinfinitefactorialhiddenmarkovmodel AT vangaelj infinitefactorialhiddenmarkovmodel AT tehy infinitefactorialhiddenmarkovmodel AT ghahramaniz infinitefactorialhiddenmarkovmodel |