Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
169 tokens/sec
GPT-4o
7 tokens/sec
Gemini 2.5 Pro Pro
45 tokens/sec
o3 Pro
4 tokens/sec
GPT-4.1 Pro
38 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

A Bayesian multilevel hidden Markov model with Poisson-lognormal emissions for intense longitudinal count data (2403.12561v1)

Published 19 Mar 2024 in stat.ME

Abstract: Hidden Markov models (HMMs) are probabilistic methods in which observations are seen as realizations of a latent Markov process with discrete states that switch over time. Moving beyond standard statistical tests, HMMs offer a statistical environment to optimally exploit the information present in multivariate time series, uncovering the latent dynamics that rule them. Here, we extend the Poisson HMM to the multilevel framework, accommodating variability between individuals with continuously distributed individual random effects following a lognormal distribution, and describe how to estimate the model in a fully parametric Bayesian framework. The proposed multilevel HMM enables probabilistic decoding of hidden state sequences from multivariate count time-series based on individual-specific parameters, and offers a framework to quantificate between-individual variability formally. Through a Monte Carlo study we show that the multilevel HMM outperforms the HMM for scenarios involving heterogeneity between individuals, demonstrating improved decoding accuracy and estimation performance of parameters of the emission distribution, and performs equally well when not between heterogeneity is present. Finally, we illustrate how to use our model to explore the latent dynamics governing complex multivariate count data in an empirical application concerning pilot whale diving behaviour in the wild, and how to identify neural states from multi-electrode recordings of motor neural cortex activity in a macaque monkey in an experimental set up. We make the multilevel HMM introduced in this study publicly available in the R-package mHMMbayes in CRAN.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (64)
  1. mHMMbayes: Multilevel Hidden Markov Models Using Bayesian Estimation. URL: https://cran.r-project.org/web/packages/mHMMbayes/mHMMbayes.pdf.
  2. Mixed Hidden Markov models: An extension of the Hidden Markov model to the longitudinal data setting. Journal of the American Statistical Association 102, 201–210. doi:10.1198/016214506000001086.
  3. BAYESIAN INFERENCE FOR A COVARIANCE MATRIX. Conference on Applied Statistics in Agriculture URL: https://newprairiepress.org/agstatconference/2014/proceedings/8, doi:10.4148/2475-7772.1004.
  4. Dynamic Structural Equation Models. Structural Equation Modeling: A Multidisciplinary Journal 25, 359–388. URL: https://www.tandfonline.com/doi/full/10.1080/10705511.2017.1406803, doi:10.1080/10705511.2017.1406803. publisher: Routledge.
  5. Scanning electron microscopy of chronically implanted intracortical microelectrode arrays in non-human primates. Journal of Neural Engineering 13, 026003. URL: https://dx.doi.org/10.1088/1741-2560/13/2/026003, doi:10.1088/1741-2560/13/2/026003. publisher: IOP Publishing.
  6. A Discrete Time Event-History Approach to Informative Drop-Out in Mixed Latent Markov Models with Covariates. Biometrics 71, 80–89. doi:10.1111/biom.12224.
  7. A Maximization Technique Occurring in the Statistical Analysis of Probabilistic Functions of Markov Chains. The Annals of Mathematical Statistics 41, 164–171. URL: https://www.jstor.org/stable/2239727. publisher: Institute of Mathematical Statistics.
  8. Classifying grey seal behaviour in relation to environmental variability and commercial fishing activity - a multivariate hidden Markov model. Scientific Reports 9, 5642. URL: https://www.nature.com/articles/s41598-019-42109-w, doi:10.1038/s41598-019-42109-w. number: 1 Publisher: Nature Publishing Group.
  9. An application of upscaled optimal foraging theory using hidden Markov modelling: year-round behavioural variation in a large arctic herbivore. Movement Ecology 8, 25. URL: https://doi.org/10.1186/s40462-020-00213-x, doi:10.1186/s40462-020-00213-x.
  10. Comparison of Multivariate Poisson lognormal spatial and temporal crash models to identify hot spots of intersections based on crash types. Accident Analysis and Prevention 99, 330–341. doi:10.1016/j.aap.2016.11.022. publisher: Elsevier Ltd.
  11. Epilepsy as a dynamic disease: A Bayesian model for differentiating seizure risk from natural variability. Epilepsia Open 3, 236–246. URL: http://doi.wiley.com/10.1002/epi4.12112, doi:10.1002/epi4.12112.
  12. Bayesian hierarchical Poisson models with a hidden Markov structure for the detection of influenza epidemic outbreaks. Statistical Methods in Medical Research 24, 206–223. URL: https://doi.org/10.1177/0962280211414853, doi:10.1177/0962280211414853. publisher: SAGE Publications Ltd STM.
  13. Count Time Series: A Methodological Review. Journal of the American Statistical Association 116, 1533–1547. URL: https://doi.org/10.1080/01621459.2021.1904957, doi:10.1080/01621459.2021.1904957. publisher: Taylor & Francis _eprint: https://doi.org/10.1080/01621459.2021.1904957.
  14. Neuroplasticity Subserving Motor Skill Learning. Neuron 72, 443–454. URL: https://www.sciencedirect.com/science/article/pii/S0896627311009184, doi:10.1016/j.neuron.2011.10.008.
  15. Maximum Likelihood from Incomplete Data Via the EM Algorithm. Journal of the Royal Statistical Society: Series B (Methodological) 39, 1–22. URL: https://onlinelibrary.wiley.com/doi/abs/10.1111/j.2517-6161.1977.tb01600.x, doi:10.1111/j.2517-6161.1977.tb01600.x. _eprint: https://onlinelibrary.wiley.com/doi/pdf/10.1111/j.2517-6161.1977.tb01600.x.
  16. A multivariate mixed hidden Markov model for blue whale behaviour and responses to sound exposure. The Annals of Applied Statistics 11, 362–392. URL: http://projecteuclid.org/euclid.aoas/1491616885, doi:10.1214/16-AOAS1008. publisher: Institute of Mathematical Statistics.
  17. Hidden Markov models for zero-inflated Poisson counts with an application to substance use. Statistics in Medicine 30, 1678–1694. URL: https://onlinelibrary.wiley.com/doi/abs/10.1002/sim.4207, doi:10.1002/sim.4207. _eprint: https://onlinelibrary.wiley.com/doi/pdf/10.1002/sim.4207.
  18. Poisson hidden markov model on earthquake occurrences in Metro Manila, Philippines. Earth Science Informatics 15, 1635–1645. URL: https://doi.org/10.1007/s12145-022-00810-x, doi:10.1007/s12145-022-00810-x.
  19. The Viterbi Algorithm. Proceedings of the IEEE 61, 268–278. doi:10.1109/PROC.1973.9030.
  20. Beyond Power Calculations: Assessing Type S (Sign) and Type M (Magnitude) Errors. Perspectives on Psychological Science 9, 641–651. URL: https://doi.org/10.1177/1745691614551642, doi:10.1177/1745691614551642.
  21. Bayesian Measures of Explained Variance and Pooling in Multilevel (Hierarchical) Models. Technometrics 48, 241–251. URL: http://www.tandfonline.com/doi/abs/10.1198/004017005000000517, doi:10.1198/004017005000000517.
  22. On the Use of Mixed Markov Models for Intensive Longitudinal Data. Multivariate Behavioral Research 52, 747–767. doi:10.1080/00273171.2017.1370364. publisher: Routledge.
  23. Hidden Markov model detection of interpersonal interaction dynamics in predicting patient depression improvement in psychotherapy: Proof-of-concept study. Journal of Affective Disorders Reports 14, 100635. URL: https://www.sciencedirect.com/science/article/pii/S2666915323001737, doi:10.1016/j.jadr.2023.100635.
  24. Bayesian nonhomogeneous Markov models via Pólya-Gamma data augmentation with applications to rainfall modeling. Annals of Applied Statistics 11, 393–426. doi:10.1214/16-AOAS1009. arXiv: 1701.02856.
  25. Multilevel analysis: techniques and applications. 3rd ed., Routledge/Taylor & Francis Group.
  26. Mixed Hidden Markov Models for Clinical Research with Discrete Repeated Measurements. American Journal of Theoretical and Applied Statistics 6, 290. URL: http://www.sciencepublishinggroup.com/journal/paperinfo?journalid=146&doi=10.11648/j.ajtas.20170606.15, doi:10.11648/j.ajtas.20170606.15.
  27. Individual, ecological, and anthropogenic influences on activity budgets of long-finned pilot whales. Ecosphere 8, e02044. URL: https://onlinelibrary.wiley.com/doi/abs/10.1002/ecs2.2044, doi:10.1002/ecs2.2044. _eprint: https://esajournals.onlinelibrary.wiley.com/doi/pdf/10.1002/ecs2.2044.
  28. A two-state mixed hidden Markov model for risky teenage driving behavior. Annals of Applied Statistics 9, 849–865. URL: http://www.vtti.vt.edu., doi:10.1214/14-AOAS765.
  29. Joint estimation over multiple individuals improves behavioural state inference from animal movement data. Scientific Reports 6, 20625. URL: https://www.nature.com/articles/srep20625, doi:10.1038/srep20625.
  30. Top-down coordination of local cortical state during selective attention. Neuron 109, 894–904.e8. URL: https://www.sciencedirect.com/science/article/pii/S0896627320309958, doi:10.1016/j.neuron.2020.12.013.
  31. Bayesian multilevel hidden Markov models identify stable state dynamics in longitudinal recordings from macaque primary motor cortex. preprint. Neuroscience. URL: http://biorxiv.org/lookup/doi/10.1101/2022.10.17.512024, doi:10.1101/2022.10.17.512024.
  32. Sample size and power calculations for medical studies by simulation when closed form expressions are not available. Statistical Methods in Medical Research 22, 324–345. doi:10.1177/0962280212439578.
  33. Flexible and practical modeling of animal telemetry data: hidden Markov models and extensions. Ecology 93, 2336–2342. URL: https://onlinelibrary.wiley.com/doi/abs/10.1890/11-2241.1, doi:10.1890/11-2241.1. _eprint: https://esajournals.onlinelibrary.wiley.com/doi/pdf/10.1890/11-2241.1.
  34. Moving beyond noninformative priors: why and how to choose weakly informative priors in Bayesian analyses. Oikos 128, 912–928. URL: https://onlinelibrary.wiley.com/doi/abs/10.1111/oik.05985, doi:10.1111/oik.05985. publisher: Blackwell Publishing Ltd.
  35. An Introduction to Animal Movement Modeling with Hidden Markov Models using Stan for Bayesian Inference URL: http://arxiv.org/abs/1806.10639. arXiv: 1806.10639.
  36. Predicting Patient-ventilator Asynchronies with Hidden Markov Models. Scientific Reports 8, 17614. URL: https://www.nature.com/articles/s41598-018-36011-0, doi:10.1038/s41598-018-36011-0. number: 1 Publisher: Nature Publishing Group.
  37. Multilevel Hidden Markov Models for Behavioral Data: A Hawk-and-Dove Experiment. Multivariate Behavioral Research 0, 1–15. URL: https://doi.org/10.1080/00273171.2021.1912583, doi:10.1080/00273171.2021.1912583. publisher: Routledge _eprint: https://doi.org/10.1080/00273171.2021.1912583.
  38. A semiparametric approach to hidden Markov models under longitudinal observations. Statistics and Computing 19, 381–393. URL: https://link-springer-com.proxy.library.uu.nl/article/10.1007/s11222-008-9099-2, doi:10.1007/s11222-008-9099-2. publisher: Springer.
  39. Worth the effort? A practical examination of random effects in hidden Markov models for animal telemetry data. Methods in Ecology and Evolution 12, 1475–1497. URL: https://onlinelibrary.wiley.com/doi/abs/10.1111/2041-210X.13619, doi:10.1111/2041-210X.13619. _eprint: https://besjournals.onlinelibrary.wiley.com/doi/pdf/10.1111/2041-210X.13619.
  40. momentuHMM: R package for generalized hidden Markov models of animal movement. Methods in Ecology and Evolution 9, 1518–1530. URL: https://onlinelibrary.wiley.com/doi/abs/10.1111/2041-210X.12995, doi:10.1111/2041-210X.12995. _eprint: https://onlinelibrary.wiley.com/doi/pdf/10.1111/2041-210X.12995.
  41. Poisson Multilevel Models with Small Samples. Multivariate Behavioral Research 54, 444–455. URL: https://www.tandfonline.com/doi/full/10.1080/00273171.2018.1545630, doi:10.1080/00273171.2018.1545630. publisher: Routledge.
  42. Go Multivariate: Recommendations on Bayesian Multilevel Hidden Markov Models with Categorical Data. Multivariate Behavioral Research 0, 1–29. URL: https://doi.org/10.1080/00273171.2023.2205392, doi:10.1080/00273171.2023.2205392. publisher: Routledge _eprint: https://doi.org/10.1080/00273171.2023.2205392.
  43. Evidence for severe mood instability in patients with bipolar disorder: Applying multilevel hidden Markov modelling to intensive longitudinal ecological momentary assessment data. URL: https://psyarxiv.com/egp82/, doi:10.31234/osf.io/egp82.
  44. Identifying Seismicity Levels via Poisson Hidden Markov Models. Pure and Applied Geophysics 167, 919–931. URL: https://doi.org/10.1007/s00024-010-0088-y, doi:10.1007/s00024-010-0088-y.
  45. Hidden Markov Models Reveal Tactical Adjustment of Temporally Clustered Courtship Displays in Response to the Behaviors of a Robotic Female URL: https://dx.doi.org/10.5061/dryad.sn0c503., doi:10.5061/dryad.sn0c503. iSBN: 201923:03:13.
  46. Sex-specific variation in the use of vertical habitat by a resident Antarctic top predator. Proceedings of the Royal Society B: Biological Sciences 287, 20201447. URL: https://royalsocietypublishing.org/doi/full/10.1098/rspb.2020.1447, doi:10.1098/rspb.2020.1447. publisher: Royal Society.
  47. Identification and initial validation of empirically derived bipolar symptom states from a large longitudinal dataset: an application of hidden Markov modeling to the Systematic Treatment Enhancement Program for Bipolar Disorder (STEP-BD) study. Psychological Medicine 49, 1102–1108. URL: https://www.cambridge.org/core/product/identifier/S0033291718002143/type/journal_article, doi:10.1017/S0033291718002143.
  48. R Core Team, 2021. R Development Core Team. R: A Language and Environment for Statistical Computing doi:http://www.R-project.org. iSBN: 3_900051_00_3.
  49. Analysis, classification, and coding of multielectrode spike trains with hidden Markov models. Biological Cybernetics 71, 359–373. URL: https://doi.org/10.1007/BF00239623, doi:10.1007/BF00239623.
  50. A Poisson Hidden Markov Model for Multiview Video Traffic. IEEE/ACM Transactions on Networking 23, 547–558. doi:10.1109/TNET.2014.2303162. conference Name: IEEE/ACM Transactions on Networking.
  51. Hierarchical Models for Heterogeneous Units, in: Bayesian Statistics and Marketing. John Wiley & Sons, Ltd, pp. 129–158. URL: https://onlinelibrary.wiley.com/doi/abs/10.1002/0470863692.ch5, doi:10.1002/0470863692.ch5. section: 5 _eprint: https://onlinelibrary.wiley.com/doi/pdf/10.1002/0470863692.ch5.
  52. A Bayesian HMM with random effects and an unknown number of states for DNA copy number analysis. Journal of Statistical Computation and Simulation 83, 82–96. doi:10.1080/00949655.2011.609818. publisher: Taylor and Francis Ltd.
  53. The Behavioral Relevance of Cortical Neural Ensemble Responses Emerges Suddenly. The Journal of Neuroscience 36, 655–669. URL: https://www.jneurosci.org/lookup/doi/10.1523/JNEUROSCI.2265-15.2016, doi:10.1523/JNEUROSCI.2265-15.2016.
  54. Towards the neural population doctrine. Current Opinion in Neurobiology 55, 103–111. URL: https://www.sciencedirect.com/science/article/pii/S0959438818300990, doi:10.1016/j.conb.2019.02.002.
  55. On the application of mixed hidden markov models to multiple behavioural time series. Interface Focus 2, 180–189. doi:10.1098/rsfs.2011.0077. publisher: Royal Society.
  56. Bayesian methods for hidden Markov models: Recursive computing in the 21st century. Journal of the American Statistical Association 97, 337–351. doi:10.1198/016214502753479464.
  57. Hidden Markov models for alcoholism treatment trial data. Annals of Applied Statistics 6, 366–395. doi:10.1214/09-AOAS282.
  58. Fluctuations in behavior and affect in college students measured using deep phenotyping. Scientific Reports 12, 1932. URL: https://www.nature.com/articles/s41598-022-05331-7, doi:10.1038/s41598-022-05331-7. number: 1 Publisher: Nature Publishing Group.
  59. Seven things to remember about hidden Markov models: A tutorial on Markovian models for time series. Journal of Mathematical Psychology 55, 403–415. doi:10.1016/j.jmp.2011.08.002.
  60. Error Bounds for Convolutional Codes and an Asymptotically Optimum Decoding Algorithm. IEEE Transactions on Information Theory 13, 260–269. doi:10.1109/TIT.1967.1054010.
  61. Ultra-small carbon fiber electrode recording site optimization and improved in vivo chronic recording yield. Journal of Neural Engineering 17, 026037. URL: https://dx.doi.org/10.1088/1741-2552/ab8343, doi:10.1088/1741-2552/ab8343. publisher: IOP Publishing.
  62. Is adding more indicators to a latent class analysis beneficial or detrimental? Results of a Monte-Carlo study. Frontiers in Psychology 5. URL: https://www.frontiersin.org/articles/10.3389/fpsyg.2014.00920.
  63. A Study of Individual Characteristics of Driving Behavior Based on Hidden Markov Model. Technical Report. URL: http://www.sensorsportal.com. publication Title: Sensors & Transducers Volume: 167.
  64. Hidden Markov models for time series: An introduction using R, second edition. Chapman and Hall/CRC. doi:10.1201/b20790. publication Title: Hidden Markov Models for Time Series: An Introduction Using R, Second Edition.

Summary

We haven't generated a summary for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize Now
X Twitter Logo Streamline Icon: https://streamlinehq.com

Tweets