{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,7,10]],"date-time":"2026-07-10T18:15:36Z","timestamp":1783707336812,"version":"3.55.0"},"reference-count":43,"publisher":"Public Library of Science (PLoS)","issue":"7","license":[{"start":{"date-parts":[[2026,7,10]],"date-time":"2026-07-10T00:00:00Z","timestamp":1783641600000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/publicdomain\/zero\/1.0\/"}],"funder":[{"name":"New Frontiers in Research Fund \u2013 Special Call","award":["NFRFR-2022-00395"],"award-info":[{"award-number":["NFRFR-2022-00395"]}]}],"content-domain":{"domain":["www.plosone.org"],"crossmark-restriction":false},"short-container-title":["PLoS One"],"abstract":"<jats:p>\n                    A Bayesian computational framework for parsimonious inference in stochastic nonlinear dynamical systems is presented. This framework enables the concurrent estimation of system states, time-varying parameters, time-invariant parameters, and the optimal sparsity structure of the model parameters. Because differential equation-based models are often simplified mechanistic or phenomenological representations, robust inference from noisy measurement data requires explicit treatment of model error and uncertainty. Model error and time-varying parameters can be represented as random processes, enabling inference while making minimal assumptions about the underlying sources of discrepancy and variability. Adopting stochastic differential equation representations affords the model significant flexibility, but can also render it susceptible to overfitting during statistical inversion, where the inferred model may track noise rather than the underlying signal. To alleviate the effects of overfitting and to enable the discovery of the optimal sparse representation of the time-invariant parameters, a Bayesian sparse learning algorithm is embedded within the framework. This sparse learning framework adopts an approximate hierarchical Bayesian setting defined by a series of semi-analytical expressions. The model structure inference framework is validated using a stochastic compartmental model for tracking and forecasting active cases of an infectious disease. Compartmental models describe population-level infectious disease dynamics through interactions among population fractions grouped by disease state. Mathematically, such models consist of a system of coupled ordinary differential equations. This example adopts an expressive compartmental model that includes multiple possible interactions between disease states, motivated by early uncertainty surrounding COVID-19 reinfection dynamics and their implications for long-term epidemic forecasting. The sparse learning exercise permits the inference of\n                    <jats:italic>a priori<\/jats:italic>\n                    unknown epidemiological dynamics from simulated public health data, discovering the nested compartmental model that optimizes the trade-off between average data-fit and model complexity. It is shown that inducing sparsity among the model parameters eliminates redundant interactions between compartments, equivalently revealing the optimal coupling structure between differential equations.\n                  <\/jats:p>","DOI":"10.1371\/journal.pone.0350747","type":"journal-article","created":{"date-parts":[[2026,7,10]],"date-time":"2026-07-10T17:38:46Z","timestamp":1783705126000},"page":"e0350747","update-policy":"https:\/\/doi.org\/10.1371\/journal.pone.corrections_policy","source":"Crossref","is-referenced-by-count":0,"title":["Semi-analytical hierarchical Bayesian inference of nonlinear model structure in stochastic dynamics: Applied to compartmental models of infectious diseases"],"prefix":"10.1371","volume":"21","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-9320-6619","authenticated-orcid":true,"given":"Brandon","family":"Robinson","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Philippe","family":"Bisaillon","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Rimple","family":"Sandhu","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Mohammad","family":"Khalil","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jodi D.","family":"Edwards","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Tetyana","family":"Kendzerska","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Thomas","family":"Walker","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Shirley","family":"Mills","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Chris","family":"Pettit","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Dominique","family":"Poirel","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-8427-8901","authenticated-orcid":true,"given":"Abhijit","family":"Sarkar","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"340","published-online":{"date-parts":[[2026,7,10]]},"reference":[{"issue":"15","key":"pone.0350747.ref001","doi-asserted-by":"crossref","first-page":"3932","DOI":"10.1073\/pnas.1517384113","article-title":"Discovering governing equations from data by sparse identification of nonlinear dynamical systems","volume":"113","author":"SL Brunton","year":"2016","journal-title":"Proc Natl Acad Sci U S A"},{"issue":"2204","key":"pone.0350747.ref002","first-page":"20170009","article-title":"Model selection for dynamical systems via sparse regression and information criteria","volume":"473","author":"NM Mangan","year":"2017","journal-title":"Proc Math Phys Eng Sci"},{"issue":"2219","key":"pone.0350747.ref003","first-page":"20180335","article-title":"Sparse identification of nonlinear dynamics for model predictive control in the low-data limit","volume":"474","author":"E Kaiser","year":"2018","journal-title":"Proc Math Phys Eng Sci"},{"key":"pone.0350747.ref004","unstructured":"Horrocks JH. Sparse identification of epidemiological models from empirical data [Master\u2019s thesis] University of Waterloo; 2018."},{"issue":"1","key":"pone.0350747.ref005","doi-asserted-by":"crossref","first-page":"7061","DOI":"10.1038\/s41598-020-63877-w","article-title":"Algorithmic discovery of dynamic models from infectious disease data","volume":"10","author":"J Horrocks","year":"2020","journal-title":"Sci Rep"},{"issue":"1","key":"pone.0350747.ref006","doi-asserted-by":"crossref","first-page":"211823","DOI":"10.1098\/rsos.211823","article-title":"Sparsifying priors for Bayesian uncertainty quantification in model discovery","volume":"9","author":"SM Hirsh","year":"2022","journal-title":"R Soc Open Sci"},{"issue":"24","key":"pone.0350747.ref007","doi-asserted-by":"crossref","first-page":"241723","DOI":"10.1063\/1.5018409","article-title":"Sparse learning of stochastic dynamical equations","volume":"148","author":"L Boninsegna","year":"2018","journal-title":"J Chem Phys"},{"key":"pone.0350747.ref008","unstructured":"Sandhu R. Model comparison and sparse learning of nonlinear physics-based models using Bayesian inference [PhD thesis] Carleton University; 2020."},{"key":"pone.0350747.ref009","doi-asserted-by":"crossref","first-page":"109728","DOI":"10.1016\/j.jcp.2020.109728","article-title":"Nonlinear sparse Bayesian learning for physics-based models","volume":"426","author":"R Sandhu","year":"2021","journal-title":"J Comput Phys"},{"key":"pone.0350747.ref010","first-page":"211","article-title":"Sparse Bayesian learning and the relevance vector machine","volume":"1","author":"ME Tipping","year":"2001","journal-title":"J Mach Learn Res"},{"key":"pone.0350747.ref011","doi-asserted-by":"crossref","first-page":"117816","DOI":"10.1016\/j.jsv.2023.117816","article-title":"Encoding nonlinear and unsteady aerodynamics of limit cycle oscillations using nonlinear sparse Bayesian learning","volume":"569","author":"R Sandhu","year":"2024","journal-title":"J Sound Vib"},{"key":"pone.0350747.ref012","unstructured":"Dabiran N, Robinson B, Sandhu R, Khalil M, Poirel D, Sarkar A. Sparse Bayesian neural networks for regression: tackling overfitting and computational challenges in uncertainty quantification. arXiv:2310.15614 [Preprint]. 2023. https:\/\/arxiv.org\/abs\/2310.15614"},{"key":"pone.0350747.ref013","unstructured":"Dabiran N, Robinson B, Sandhu R, Khalil M, Pettit CL, Poirel D, et al. Exploring hierarchical framework of nonlinear sparse Bayesian learning algorithm through numerical investigations. arXiv:2310.14749 [Preprint]. 2023. https:\/\/arxiv.org\/abs\/2310.14749"},{"key":"pone.0350747.ref014","doi-asserted-by":"crossref","first-page":"81","DOI":"10.1016\/j.jsv.2014.10.002","article-title":"The estimation of time-invariant parameters of noisy nonlinear oscillatory systems","volume":"344","author":"M Khalil","year":"2015","journal-title":"J Sound Vib"},{"key":"pone.0350747.ref015","doi-asserted-by":"crossref","first-page":"118106","DOI":"10.1016\/j.jsv.2023.118106","article-title":"Robust Bayesian state and parameter estimation framework for stochastic dynamical systems with combined time-varying and time-invariant parameters","volume":"575","author":"P Bisaillon","year":"2024","journal-title":"J Sound Vib"},{"key":"pone.0350747.ref016","article-title":"A Bayesian model calibration framework for stochastic compartmental models with both time-varying and time-invariant parameters","author":"B Robinson","year":"2024","journal-title":"Infect Dis Model"},{"issue":"7","key":"pone.0350747.ref017","doi-asserted-by":"crossref","first-page":"1377","DOI":"10.1093\/aje\/kwab013","article-title":"Quantifying uncertainty in mechanistic models of infectious disease","volume":"190","author":"L D\u2019Agostino McGowan","year":"2021","journal-title":"Am J Epidemiol"},{"key":"pone.0350747.ref018","unstructured":"Robinson B. Robust, sparse, and Bayesian learning of stochastic nonlinear dynamical systems with applications in aeroelasticity and infectious disease modelling [PhD thesis]. Carleton University; 2025."},{"issue":"3","key":"pone.0350747.ref019","doi-asserted-by":"crossref","DOI":"10.1136\/bmjopen-2021-052681","article-title":"Comprehensive compartmental model and calibration algorithm for the study of clinical implications of the population-level spread of COVID-19: a study protocol","volume":"12","author":"B Robinson","year":"2022","journal-title":"BMJ Open"},{"key":"pone.0350747.ref020","volume-title":"Machine learning: a probabilistic perspective","author":"KP Murphy","year":"2012"},{"issue":"4","key":"pone.0350747.ref021","doi-asserted-by":"crossref","first-page":"104810","DOI":"10.1016\/j.compbiomed.2021.104810","article-title":"Extended Kalman filter based on stochastic epidemiological model for COVID-19 modelling","volume":"137","author":"X Zhu","year":"2021","journal-title":"Comput Biol Med"},{"issue":"4","key":"pone.0350747.ref022","doi-asserted-by":"crossref","first-page":"1441","DOI":"10.1109\/JBHI.2021.3063106","article-title":"Monitoring and tracking the evolution of a viral epidemic through nonlinear kalman filtering: application to the COVID-19 case","volume":"26","author":"A Gomez-Exposito","year":"2022","journal-title":"IEEE J Biomed Health Inform"},{"issue":"7","key":"pone.0350747.ref023","doi-asserted-by":"crossref","DOI":"10.1371\/journal.pgph.0003467","article-title":"Bayesian model selection for COVID-19 pandemic state estimation using extended Kalman filters: case study for Saudi Arabia","volume":"4","author":"L Alyami","year":"2024","journal-title":"PLOS Glob Public Health"},{"issue":"8","key":"pone.0350747.ref024","article-title":"An application of the ensemble Kalman filter in epidemiological modelling","volume":"16","author":"R Lal","year":"2021","journal-title":"PLoS One"},{"issue":"3","key":"pone.0350747.ref025","doi-asserted-by":"crossref","first-page":"413","DOI":"10.3934\/fods.2021001","article-title":"An international initiative of predicting the SARS-CoV-2 pandemic using ensemble data assimilation","volume":"3","author":"G Evensen","year":"2021","journal-title":"Found Data Sci"},{"issue":"1","key":"pone.0350747.ref026","doi-asserted-by":"crossref","first-page":"2661","DOI":"10.1038\/s41598-019-38930-y","article-title":"Real-time forecasting of hand-foot-and-mouth disease outbreaks using the integrating compartment model and assimilation filtering","volume":"9","author":"Z Zhan","year":"2019","journal-title":"Sci Rep"},{"issue":"1","key":"pone.0350747.ref027","doi-asserted-by":"crossref","DOI":"10.1371\/journal.pcbi.1005257","article-title":"A likelihood approach for real-time calibration of stochastic compartmental epidemic models","volume":"13","author":"C Zimmer","year":"2017","journal-title":"PLoS Comput Biol"},{"issue":"11","key":"pone.0350747.ref028","article-title":"Applying particle filtering in both aggregated and age-structured population compartmental models of pre-vaccination measles","volume":"13","author":"X Li","year":"2018","journal-title":"PLoS One"},{"issue":"11","key":"pone.0350747.ref029","doi-asserted-by":"crossref","first-page":"115008","DOI":"10.1088\/1361-6420\/ac2cdc","article-title":"Bayesian particle filter algorithm for learning epidemic dynamics","volume":"37","author":"D Calvetti","year":"2021","journal-title":"Inverse Probl"},{"key":"pone.0350747.ref030","doi-asserted-by":"crossref","first-page":"21","DOI":"10.1016\/j.mbs.2014.06.018","article-title":"Comparison of the performance of particle filter algorithms applied to tracking of a disease epidemic","volume":"255","author":"DM Sheinson","year":"2014","journal-title":"Math Biosci"},{"key":"pone.0350747.ref031","first-page":"473","article-title":"Practical Markov chain Monte Carlo","author":"CJ Geyer","year":"1992","journal-title":"Stat Sci"},{"key":"pone.0350747.ref032","doi-asserted-by":"crossref","DOI":"10.1201\/b14835","volume-title":"Markov chain Monte Carlo in practice","author":"WR Gilks","year":"1995"},{"issue":"518","key":"pone.0350747.ref033","doi-asserted-by":"crossref","first-page":"859","DOI":"10.1080\/01621459.2017.1285773","article-title":"Variational inference: a review for statisticians","volume":"112","author":"DM Blei","year":"2017","journal-title":"J Am Stat Assoc"},{"issue":"1","key":"pone.0350747.ref034","doi-asserted-by":"crossref","first-page":"25","DOI":"10.1023\/A:1008932416310","article-title":"Bayesian parameter estimation via variational methods","volume":"10","author":"TS Jaakkola","year":"2000","journal-title":"Stat Comput"},{"key":"pone.0350747.ref035","doi-asserted-by":"crossref","first-page":"103540","DOI":"10.1016\/j.probengmech.2023.103540","article-title":"Robust scalable initialization for Bayesian variational inference with multi-modal Laplace approximations","volume":"74","author":"W Bridgman","year":"2023","journal-title":"Probabilistic Eng Mech"},{"key":"pone.0350747.ref036","doi-asserted-by":"crossref","DOI":"10.1007\/978-3-642-03711-5","volume-title":"Data assimilation: the ensemble Kalman filter","author":"G Evensen","year":"2009"},{"issue":"3","key":"pone.0350747.ref037","doi-asserted-by":"crossref","first-page":"448","DOI":"10.1162\/neco.1992.4.3.448","article-title":"A practical Bayesian framework for backpropagation networks","volume":"4","author":"DCJ MacKay","year":"1992","journal-title":"Neural Comput"},{"issue":"6","key":"pone.0350747.ref038","doi-asserted-by":"crossref","first-page":"351","DOI":"10.1038\/s41577-020-0316-3","article-title":"Will we see protection or reinfection in COVID-19?","volume":"20","author":"O Miyo","year":"2020","journal-title":"Nat Rev Immunol"},{"issue":"772","key":"pone.0350747.ref039","doi-asserted-by":"crossref","first-page":"700","DOI":"10.1098\/rspa.1927.0118","article-title":"A contribution to the mathematical theory of epidemics","volume":"115","author":"WO Kermack","year":"1927","journal-title":"Proc R Soc Lond A"},{"key":"pone.0350747.ref040","volume-title":"Mathematical epidemiology, vol. 1945","author":"F Brauer","year":"2008"},{"key":"pone.0350747.ref041","volume-title":"Recursive Bayesian inference on stochastic differential equations","author":"S S\u00e4rkk\u00e4","year":"2006"},{"key":"pone.0350747.ref042","doi-asserted-by":"crossref","first-page":"460","DOI":"10.1016\/j.arcontrol.2021.03.008","article-title":"Forecasting COVID-19 cases based on a parameter-varying stochastic SIR model","volume":"51","author":"JP Hespanha","year":"2021","journal-title":"Annu Rev Control"},{"issue":"27","key":"pone.0350747.ref043","doi-asserted-by":"crossref","first-page":"6209","DOI":"10.1002\/sim.9164","article-title":"Bayesian workflow for disease transmission modeling in Stan","volume":"40","author":"L Grinsztajn","year":"2021","journal-title":"Stat Med"}],"container-title":["PLOS One"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pone.0350747","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,7,10]],"date-time":"2026-07-10T17:38:55Z","timestamp":1783705135000},"score":1,"resource":{"primary":{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pone.0350747"}},"subtitle":[],"editor":[{"given":"Xuebo","family":"Zhang","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"editor"}]}],"short-title":[],"issued":{"date-parts":[[2026,7,10]]},"references-count":43,"journal-issue":{"issue":"7","published-online":{"date-parts":[[2026,7,10]]}},"URL":"https:\/\/doi.org\/10.1371\/journal.pone.0350747","relation":{},"ISSN":["1932-6203"],"issn-type":[{"value":"1932-6203","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,7,10]]}}}