{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2025,7,20]],"date-time":"2025-07-20T03:46:25Z","timestamp":1752983185944},"reference-count":37,"publisher":"MIT Press","issue":"1","content-domain":{"domain":["direct.mit.edu"],"crossmark-restriction":true},"short-container-title":[],"published-print":{"date-parts":[[2023,1,1]]},"abstract":"<jats:title>Abstract<\/jats:title><jats:p>We propose a nonlinear probabilistic generative model of Koopman mode decomposition based on an unsupervised gaussian process. Existing data-driven methods for Koopman mode decomposition have focused on estimating the quantities specified by Koopman mode decomposition: eigenvalues, eigenfunctions, and modes. Our model enables the simultaneous estimation of these quantities and latent variables governed by an unknown dynamical system. Furthermore, we introduce an efficient strategy to estimate the parameters of our model by low-rank approximations of covariance matrices. Applying the proposed model to both synthetic data and a real-world epidemiological data set, we show that various analyses are available using the estimated parameters.<\/jats:p>","DOI":"10.1162\/neco_a_01555","type":"journal-article","created":{"date-parts":[[2022,11,24]],"date-time":"2022-11-24T14:00:45Z","timestamp":1669298445000},"page":"82-103","update-policy":"http:\/\/dx.doi.org\/10.1162\/mitpressjournals.corrections.policy","source":"Crossref","is-referenced-by-count":6,"title":["Gaussian Process Koopman Mode Decomposition"],"prefix":"10.1162","volume":"35","author":[{"given":"Takahiro","family":"Kawashima","sequence":"first","affiliation":[{"name":"Department of Statistical Science, Graduate University of Advanced Studies, Tokyo 190-0014, Japan tkawa@ism.ac.jp"}]},{"given":"Hideitsu","family":"Hino","sequence":"additional","affiliation":[{"name":"Department of Statistical Science, Graduate University of Advanced Studies, Tokyo 190-0014, Japan"},{"name":"Institute of Statistical Mathematics, Tokyo 190-0014, Japan"},{"name":"RIKEN AIP, Tokyo 103-0027, Japan hino@ism.ac.jp"}]}],"member":"281","published-online":{"date-parts":[[2023,1,1]]},"reference":[{"key":"2023032023361053000_","article-title":"Multi-task gaussian process prediction","volume-title":"Advances in neural information processing systems","author":"Bonilla","year":"2007"},{"issue":"5","key":"2023032023361053000_","doi-asserted-by":"crossref","first-page":"2007","DOI":"10.1007\/s00332-019-09582-z","article-title":"Koopman operator spectrum for random dynamical systems","volume":"30","author":"\u010crnjari\u0107-\u017dic","year":"2020","journal-title":"Journal of Nonlinear Science"},{"issue":"3","key":"2023032023361053000_","doi-asserted-by":"publisher","DOI":"10.1007\/s00348-016-2127-7","article-title":"Characterizing and correcting for the effect of sensor noise in the dynamic mode decomposition","volume":"57","author":"Dawson","year":"2016","journal-title":"Experiments in Fluids"},{"issue":"2","key":"2023032023361053000_","doi-asserted-by":"publisher","first-page":"97","DOI":"10.1111\/j.1600-0870.1986.tb00459.x","article-title":"Variational algorithms for analysis and assimilation of meteorological observations: Theoretical aspects","volume":"38A","author":"Dimet","year":"1986","journal-title":"Tellus A"},{"issue":"72","key":"2023032023361053000_","first-page":"2153","article-title":"On the Nystr\u00f6m method for approximating a gram matrix for improved kernel-based learning","volume":"6","author":"Drineas","year":"2005","journal-title":"Journal of Machine Learning Research"},{"issue":"4","key":"2023032023361053000_","doi-asserted-by":"publisher","first-page":"343","DOI":"10.1007\/s10236-003-0036-9","article-title":"The ensemble Kalman filter: Theoretical formulation and practical implementation","volume":"53","author":"Evensen","year":"2003","journal-title":"Ocean Dynamics"},{"issue":"2","key":"2023032023361053000_","doi-asserted-by":"publisher","DOI":"10.1049\/ip-f-2.1993.0015","article-title":"Novel approach to nonlinear\/non-gaussian Bayesian state estimation","volume":"140","author":"Gordon","year":"1993","journal-title":"IEE Proceedings F Radar and Signal Processing"},{"journal-title":"Low-rank dynamic mode decomposition: Optimal solution in polynomial-time","year":"2020","author":"H\u00e9as","key":"2023032023361053000_"},{"issue":"2","key":"2023032023361053000_","doi-asserted-by":"crossref","DOI":"10.1063\/1.4863670","article-title":"Sparsity-promoting dynamic mode decomposition","volume":"26","author":"Jovanovi\u0107","year":"2014","journal-title":"Physics of Fluids"},{"issue":"Series D","key":"2023032023361053000_","doi-asserted-by":"publisher","first-page":"35","DOI":"10.1115\/1.3662552","article-title":"A new approach to linear filtering and prediction problems","volume":"82","author":"Kalman","year":"1960","journal-title":"Transactions of the ASME: Journal of Basic Engineering"},{"key":"2023032023361053000_","doi-asserted-by":"crossref","DOI":"10.1007\/978-3-642-66282-9","volume-title":"Perturbation theory for linear operators","author":"Kat\u014d","year":"1995"},{"key":"2023032023361053000_","first-page":"911","article-title":"Dynamic mode decomposition with reproducing kernels for Koopman spectral analysis","volume-title":"Advances in neural information processing systems","author":"Kawahara","year":"2016"},{"issue":"9","key":"2023032023361053000_","doi-asserted-by":"publisher","first-page":"8083","DOI":"10.1609\/aaai.v35i9.16985","article-title":"Bayesian dynamic mode decomposition with variational matrix factorization","volume":"35","author":"Kawashima","year":"2021","journal-title":"Proceedings of the AAAI Conference on Artificial Intelligence"},{"issue":"1","key":"2023032023361053000_","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1080\/10618600.1996.10474692","article-title":"Monte Carlo filter and smoother for non-gaussian nonlinear state space models","volume":"5","author":"Kitagawa","year":"1996","journal-title":"Journal of Computational and Graphical Statistics"},{"issue":"60","key":"2023032023361053000_","first-page":"1783","article-title":"Probabilistic non-linear principal component analysis with gaussian process latent variable models","volume":"6","author":"Lawrence","year":"2005","journal-title":"Journal of Machine Learning Research"},{"issue":"2","key":"2023032023361053000_","doi-asserted-by":"publisher","first-page":"882","DOI":"10.1137\/15M1054924","article-title":"Higher order dynamic mode decomposition","volume":"16","author":"Le Clainche","year":"2017","journal-title":"SIAM Journal on Applied Dynamical Systems"},{"issue":"4","key":"2023032023361053000_","doi-asserted-by":"publisher","first-page":"309","DOI":"10.1111\/j.1600-0870.1985.tb00430.x","article-title":"The use of adjoint equations to solve a variational adjustment problem with advective constraints","volume":"37A","author":"Lewis","year":"1985","journal-title":"Tellus A"},{"issue":"2","key":"2023032023361053000_","doi-asserted-by":"publisher","first-page":"449","DOI":"10.1016\/j.ifacol.2020.12.217","article-title":"On gaussian process based Koopman operators","volume":"53","author":"Lian","year":"2020","journal-title":"IFAC-PapersOnLine"},{"key":"2023032023361053000_","first-page":"15","article-title":"Variational Bayesian approach to movie rating prediction","volume":"7","author":"Lim","year":"2007","journal-title":"Proceedings of KDD Cup and Workshop"},{"journal-title":"Application of gaussian process regression to Koopman mode decomposition for noisy dynamic data","year":"2019","author":"Masuda","key":"2023032023361053000_"},{"issue":"1","key":"2023032023361053000_","doi-asserted-by":"crossref","first-page":"309","DOI":"10.1007\/s11071-005-2824-x","article-title":"Spectral properties of dynamical systems, model reduction and decompositions","volume":"41","author":"Mezi\u0107","year":"2005","journal-title":"Nonlinear Dynamics"},{"issue":"79","key":"2023032023361053000_","first-page":"2583","article-title":"Theoretical analysis of Bayesian matrix factorization","volume":"12","author":"Nakajima","year":"2011","journal-title":"Journal of Machine Learning Research"},{"issue":"2","key":"2023032023361053000_","doi-asserted-by":"publisher","first-page":"139","DOI":"10.1093\/inthealth\/ihv009","article-title":"Discovering dynamic patterns from infectious disease data using dynamic mode decomposition","volume":"7","author":"Proctor","year":"2015","journal-title":"International Health"},{"key":"2023032023361053000_","article-title":"Random features for large-scale kernel machines","volume-title":"Advances in neural information processing systems","author":"Rahimi","year":"2007"},{"key":"2023032023361053000_","article-title":"It is all in the noise: Efficient multi-task gaussian process inference with structured residuals","volume-title":"Advances in neural information processing systems","author":"Rakitsch","year":"2013"},{"key":"2023032023361053000_","doi-asserted-by":"publisher","first-page":"115","DOI":"10.1017\/S0022112009992059","article-title":"Spectral analysis of nonlinear flows","volume":"641","author":"Rowley","year":"2009","journal-title":"Journal of Fluid Mechanics"},{"key":"2023032023361053000_","doi-asserted-by":"publisher","first-page":"5","DOI":"10.1017\/S0022112010001217","article-title":"Dynamic mode decomposition of numerical and experimental data","volume":"656","author":"Schmid","year":"2010","journal-title":"Journal of Fluid Mechanics"},{"key":"2023032023361053000_","article-title":"Efficient inference in matrix-variate gaussian models with iid observation noise","volume-title":"Advances in neural information processing systems","author":"Stegle","year":"2011"},{"key":"2023032023361053000_","doi-asserted-by":"crossref","first-page":"2814","DOI":"10.24963\/ijcai.2017\/392","article-title":"Bayesian dynamic mode decomposition","volume-title":"Proceedings of the Twenty-Sixth International Joint Conference on Artificial Intelligence","author":"Takeishi","year":"2017"},{"key":"2023032023361053000_","article-title":"Learning Koopman invariant subspaces for dynamic mode decomposition","volume-title":"Advances in neural information processing systems","author":"Takeishi","year":"2017"},{"key":"2023032023361053000_","first-page":"567","article-title":"Variational learning of inducing variables in sparse gaussian processes","volume-title":"Proceedings of the Twelfth International Conference on Artificial Intelligence and Statistics","author":"Titsias","year":"2009"},{"key":"2023032023361053000_","first-page":"844","article-title":"Bayesian gaussian process latent variable model","volume-title":"Proceedings of the Thirteenth International Conference on Artificial Intelligence and Statistics","author":"Titsias","year":"2010"},{"key":"2023032023361053000_","article-title":"Gaussian process dynamical models","volume-title":"Advances in neural information processing systems","author":"Wang","year":"2005"},{"volume-title":"Connections and equivalences between the Nystr\u00f6m method and sparse variational gaussian processes","year":"2021","author":"Wild","key":"2023032023361053000_"},{"issue":"6","key":"2023032023361053000_","doi-asserted-by":"publisher","first-page":"1307","DOI":"10.1007\/s00332-015-9258-5","article-title":"A data\u2013driven approximation of the Koopman operator: Extending dynamic mode decomposition","volume":"25","author":"Williams","year":"2015","journal-title":"Journal of Nonlinear Science"},{"issue":"2","key":"2023032023361053000_","doi-asserted-by":"publisher","first-page":"247","DOI":"10.3934\/jcd.2015005","article-title":"A kernel-based method for data-driven Koopman spectral analysis","volume":"2","author":"Williams","year":"2015","journal-title":"Journal of Computational Dynamics"},{"key":"2023032023361053000_","volume-title":"Evolution of a Pandemic: A(H1N1) 2009, April 2009\u2013August 2010","author":"World Health Organization","year":"2013","edition":"2nd"}],"container-title":["Neural Computation"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/direct.mit.edu\/neco\/article-pdf\/35\/1\/82\/2075426\/neco_a_01555.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/direct.mit.edu\/neco\/article-pdf\/35\/1\/82\/2075426\/neco_a_01555.pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2024,10,9]],"date-time":"2024-10-09T13:22:05Z","timestamp":1728480125000},"score":1,"resource":{"primary":{"URL":"https:\/\/direct.mit.edu\/neco\/article\/35\/1\/82\/113810\/Gaussian-Process-Koopman-Mode-Decomposition"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2023,1,1]]},"references-count":37,"journal-issue":{"issue":"1","published-online":{"date-parts":[[2023,1,1]]},"published-print":{"date-parts":[[2023,1,1]]}},"URL":"https:\/\/doi.org\/10.1162\/neco_a_01555","relation":{},"ISSN":["0899-7667","1530-888X"],"issn-type":[{"type":"print","value":"0899-7667"},{"type":"electronic","value":"1530-888X"}],"subject":[],"published-other":{"date-parts":[[2023,1]]},"published":{"date-parts":[[2023,1,1]]}}}