Variance-Reduced Stochastic Optimization for Efficient Inference of Hidden Markov Models ...
Hidden Markov models (HMMs) are popular models to identify a finite number of latent states from sequential data. However, fitting them to large data sets can be computationally demanding because most likelihood maximization techniques require iterating through the entire underlying data set for eve...
Main Authors: | , , , , , |
---|---|
Format: | Text |
Language: | unknown |
Published: |
arXiv
2023
|
Subjects: | |
Online Access: | https://dx.doi.org/10.48550/arxiv.2310.04620 https://arxiv.org/abs/2310.04620 |
id |
ftdatacite:10.48550/arxiv.2310.04620 |
---|---|
record_format |
openpolar |
spelling |
ftdatacite:10.48550/arxiv.2310.04620 2024-09-09T20:02:21+00:00 Variance-Reduced Stochastic Optimization for Efficient Inference of Hidden Markov Models ... Sidrow, Evan Heckman, Nancy Bouchard-Côté, Alexandre Fortune, Sarah M. E. Trites, Andrew W. Auger-Méthé, Marie 2023 https://dx.doi.org/10.48550/arxiv.2310.04620 https://arxiv.org/abs/2310.04620 unknown arXiv https://dx.doi.org/10.1080/10618600.2024.2350476 Creative Commons Attribution 4.0 International https://creativecommons.org/licenses/by/4.0/legalcode cc-by-4.0 Computation stat.CO FOS Computer and information sciences Article article-journal Text ScholarlyArticle 2023 ftdatacite https://doi.org/10.48550/arxiv.2310.0462010.1080/10618600.2024.2350476 2024-06-17T08:32:40Z Hidden Markov models (HMMs) are popular models to identify a finite number of latent states from sequential data. However, fitting them to large data sets can be computationally demanding because most likelihood maximization techniques require iterating through the entire underlying data set for every parameter update. We propose a novel optimization algorithm that updates the parameters of an HMM without iterating through the entire data set. Namely, we combine a partial E step with variance-reduced stochastic optimization within the M step. We prove the algorithm converges under certain regularity conditions. We test our algorithm empirically using a simulation study as well as a case study of kinematic data collected using suction-cup attached biologgers from eight northern resident killer whales (Orcinus orca) off the western coast of Canada. In both, our algorithm converges in fewer epochs and to regions of higher likelihood compared to standard numerical optimization techniques. Our algorithm allows ... : 23 pages, 7 figures. Code available at https://github.com/evsi8432/sublinear-HMM-inference ... Text Orca Orcinus orca DataCite Canada |
institution |
Open Polar |
collection |
DataCite |
op_collection_id |
ftdatacite |
language |
unknown |
topic |
Computation stat.CO FOS Computer and information sciences |
spellingShingle |
Computation stat.CO FOS Computer and information sciences Sidrow, Evan Heckman, Nancy Bouchard-Côté, Alexandre Fortune, Sarah M. E. Trites, Andrew W. Auger-Méthé, Marie Variance-Reduced Stochastic Optimization for Efficient Inference of Hidden Markov Models ... |
topic_facet |
Computation stat.CO FOS Computer and information sciences |
description |
Hidden Markov models (HMMs) are popular models to identify a finite number of latent states from sequential data. However, fitting them to large data sets can be computationally demanding because most likelihood maximization techniques require iterating through the entire underlying data set for every parameter update. We propose a novel optimization algorithm that updates the parameters of an HMM without iterating through the entire data set. Namely, we combine a partial E step with variance-reduced stochastic optimization within the M step. We prove the algorithm converges under certain regularity conditions. We test our algorithm empirically using a simulation study as well as a case study of kinematic data collected using suction-cup attached biologgers from eight northern resident killer whales (Orcinus orca) off the western coast of Canada. In both, our algorithm converges in fewer epochs and to regions of higher likelihood compared to standard numerical optimization techniques. Our algorithm allows ... : 23 pages, 7 figures. Code available at https://github.com/evsi8432/sublinear-HMM-inference ... |
format |
Text |
author |
Sidrow, Evan Heckman, Nancy Bouchard-Côté, Alexandre Fortune, Sarah M. E. Trites, Andrew W. Auger-Méthé, Marie |
author_facet |
Sidrow, Evan Heckman, Nancy Bouchard-Côté, Alexandre Fortune, Sarah M. E. Trites, Andrew W. Auger-Méthé, Marie |
author_sort |
Sidrow, Evan |
title |
Variance-Reduced Stochastic Optimization for Efficient Inference of Hidden Markov Models ... |
title_short |
Variance-Reduced Stochastic Optimization for Efficient Inference of Hidden Markov Models ... |
title_full |
Variance-Reduced Stochastic Optimization for Efficient Inference of Hidden Markov Models ... |
title_fullStr |
Variance-Reduced Stochastic Optimization for Efficient Inference of Hidden Markov Models ... |
title_full_unstemmed |
Variance-Reduced Stochastic Optimization for Efficient Inference of Hidden Markov Models ... |
title_sort |
variance-reduced stochastic optimization for efficient inference of hidden markov models ... |
publisher |
arXiv |
publishDate |
2023 |
url |
https://dx.doi.org/10.48550/arxiv.2310.04620 https://arxiv.org/abs/2310.04620 |
geographic |
Canada |
geographic_facet |
Canada |
genre |
Orca Orcinus orca |
genre_facet |
Orca Orcinus orca |
op_relation |
https://dx.doi.org/10.1080/10618600.2024.2350476 |
op_rights |
Creative Commons Attribution 4.0 International https://creativecommons.org/licenses/by/4.0/legalcode cc-by-4.0 |
op_doi |
https://doi.org/10.48550/arxiv.2310.0462010.1080/10618600.2024.2350476 |
_version_ |
1809934305873362944 |