Flexible estimation of the state dwell-time distribution in hidden semi-Markov models
Hidden semi-Markov models generalise hidden Markov models by explicitly modelling the time spent in a given state, the so-called dwell time, using some distribution defined on the natural numbers. While the (shifted) Poisson and negative binomial distribution provide natural choices for such distrib...
Published in: | Computational Statistics & Data Analysis |
---|---|
Main Authors: | , , |
Format: | Article in Journal/Newspaper |
Language: | English |
Published: |
2022
|
Subjects: | |
Online Access: | https://publishup.uni-potsdam.de/frontdoor/index/index/docId/63358 https://doi.org/10.1016/j.csda.2022.107479 |
id |
ftubpotsdam:oai:kobv.de-opus4-uni-potsdam:63358 |
---|---|
record_format |
openpolar |
spelling |
ftubpotsdam:oai:kobv.de-opus4-uni-potsdam:63358 2024-05-19T07:41:21+00:00 Flexible estimation of the state dwell-time distribution in hidden semi-Markov models Pohle, Jennifer (Dr.) Adam, Timo (Prof. Dr.) Beumer, Larissa (Dr.) 2022-08-01 https://publishup.uni-potsdam.de/frontdoor/index/index/docId/63358 https://doi.org/10.1016/j.csda.2022.107479 eng eng https://publishup.uni-potsdam.de/frontdoor/index/index/docId/63358 https://doi.org/10.1016/j.csda.2022.107479 info:eu-repo/semantics/closedAccess ddc:510 Institut für Mathematik article doc-type:article 2022 ftubpotsdam https://doi.org/10.1016/j.csda.2022.107479 2024-04-30T23:32:08Z Hidden semi-Markov models generalise hidden Markov models by explicitly modelling the time spent in a given state, the so-called dwell time, using some distribution defined on the natural numbers. While the (shifted) Poisson and negative binomial distribution provide natural choices for such distributions, in practice, parametric distributions can lack the flexibility to adequately model the dwell times. To overcome this problem, a penalised maximum likelihood approach is proposed that allows for a flexible and data-driven estimation of the dwell-time distributions without the need to make any distributional assumption. This approach is suitable for direct modelling purposes or as an exploratory tool to investigate the latent state dynamics. The feasibility and potential of the suggested approach is illustrated in a simulation study and by modelling muskox movements in northeast Greenland using GPS tracking data. The proposed method is implemented in the R-package PHSMM which is available on CRAN. Article in Journal/Newspaper Greenland muskox University of Potsdam: publish.UP Computational Statistics & Data Analysis 172 107479 |
institution |
Open Polar |
collection |
University of Potsdam: publish.UP |
op_collection_id |
ftubpotsdam |
language |
English |
topic |
ddc:510 Institut für Mathematik |
spellingShingle |
ddc:510 Institut für Mathematik Pohle, Jennifer (Dr.) Adam, Timo (Prof. Dr.) Beumer, Larissa (Dr.) Flexible estimation of the state dwell-time distribution in hidden semi-Markov models |
topic_facet |
ddc:510 Institut für Mathematik |
description |
Hidden semi-Markov models generalise hidden Markov models by explicitly modelling the time spent in a given state, the so-called dwell time, using some distribution defined on the natural numbers. While the (shifted) Poisson and negative binomial distribution provide natural choices for such distributions, in practice, parametric distributions can lack the flexibility to adequately model the dwell times. To overcome this problem, a penalised maximum likelihood approach is proposed that allows for a flexible and data-driven estimation of the dwell-time distributions without the need to make any distributional assumption. This approach is suitable for direct modelling purposes or as an exploratory tool to investigate the latent state dynamics. The feasibility and potential of the suggested approach is illustrated in a simulation study and by modelling muskox movements in northeast Greenland using GPS tracking data. The proposed method is implemented in the R-package PHSMM which is available on CRAN. |
format |
Article in Journal/Newspaper |
author |
Pohle, Jennifer (Dr.) Adam, Timo (Prof. Dr.) Beumer, Larissa (Dr.) |
author_facet |
Pohle, Jennifer (Dr.) Adam, Timo (Prof. Dr.) Beumer, Larissa (Dr.) |
author_sort |
Pohle, Jennifer (Dr.) |
title |
Flexible estimation of the state dwell-time distribution in hidden semi-Markov models |
title_short |
Flexible estimation of the state dwell-time distribution in hidden semi-Markov models |
title_full |
Flexible estimation of the state dwell-time distribution in hidden semi-Markov models |
title_fullStr |
Flexible estimation of the state dwell-time distribution in hidden semi-Markov models |
title_full_unstemmed |
Flexible estimation of the state dwell-time distribution in hidden semi-Markov models |
title_sort |
flexible estimation of the state dwell-time distribution in hidden semi-markov models |
publishDate |
2022 |
url |
https://publishup.uni-potsdam.de/frontdoor/index/index/docId/63358 https://doi.org/10.1016/j.csda.2022.107479 |
genre |
Greenland muskox |
genre_facet |
Greenland muskox |
op_relation |
https://publishup.uni-potsdam.de/frontdoor/index/index/docId/63358 https://doi.org/10.1016/j.csda.2022.107479 |
op_rights |
info:eu-repo/semantics/closedAccess |
op_doi |
https://doi.org/10.1016/j.csda.2022.107479 |
container_title |
Computational Statistics & Data Analysis |
container_volume |
172 |
container_start_page |
107479 |
_version_ |
1799480942137442304 |