{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2025,9,24]],"date-time":"2025-09-24T10:23:33Z","timestamp":1758709413364,"version":"3.40.5"},"reference-count":73,"publisher":"Walter de Gruyter GmbH","issue":"3","funder":[{"DOI":"10.13039\/501100012226","name":"Fundamental Research Funds for the Central Universities","doi-asserted-by":"publisher","award":["22120210133"],"award-info":[{"award-number":["22120210133"]}],"id":[{"id":"10.13039\/501100012226","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2021,7,1]]},"abstract":"<jats:title>Abstract<\/jats:title>\n               <jats:p>Transfer operators such as Perron\u2013Frobenius and Koopman operator play a key role in modeling and analysis of complex dynamical systems, which allow linear representations of nonlinear dynamics by transforming the original state variables to feature spaces.\nHowever, it remains challenging to identify the optimal low-dimensional feature mappings from data.\nThe variational approach for Markov processes (VAMP) provides a comprehensive framework for the evaluation and optimization of feature mappings based on the variational estimation of modeling errors, but it still suffers from a flawed assumption on the transfer operator and therefore sometimes fails to capture the essential structure of system dynamics.\nIn this paper, we develop a powerful alternative to VAMP, called kernel embedding based variational approach for dynamical systems (KVAD).\nBy using the distance measure of functions in the kernel embedding space, KVAD effectively overcomes theoretical and practical limitations of VAMP.\nIn addition, we develop a data-driven KVAD algorithm for seeking the ideal feature mapping within a subspace spanned by given basis functions, and numerical experiments show that the proposed algorithm can significantly improve the modeling accuracy compared to VAMP.<\/jats:p>","DOI":"10.1515\/cmam-2020-0130","type":"journal-article","created":{"date-parts":[[2021,6,10]],"date-time":"2021-06-10T01:42:48Z","timestamp":1623289368000},"page":"635-659","source":"Crossref","is-referenced-by-count":7,"title":["Kernel Embedding Based Variational Approach for Low-Dimensional Approximation of Dynamical Systems"],"prefix":"10.1515","volume":"21","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-9418-6306","authenticated-orcid":false,"given":"Wenchong","family":"Tian","sequence":"first","affiliation":[{"name":"College of Environmental Science and Engineering , Tongji University , Shanghai 200092 , P. R. China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-2170-0618","authenticated-orcid":false,"given":"Hao","family":"Wu","sequence":"additional","affiliation":[{"name":"School of Mathematical Sciences , Tongji University , Shanghai 200092 , P. R. China"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"374","published-online":{"date-parts":[[2021,6,9]]},"reference":[{"key":"2023033111373295703_j_cmam-2020-0130_ref_001","doi-asserted-by":"crossref","unstructured":"Z. Bai, D. Lu and B. Vandereycken,\nRobust Rayleigh quotient minimization and nonlinear eigenvalue problems,\nSIAM J. Sci. Comput. 40 (2018), no. 5, A3495\u2013A3522.","DOI":"10.1137\/18M1167681"},{"key":"2023033111373295703_j_cmam-2020-0130_ref_002","doi-asserted-by":"crossref","unstructured":"S. L. Brunton, B. W. Brunton, J. L. Proctor and J. N. Kutz,\nKoopman invariant subspaces and finite linear representations of nonlinear dynamical systems for control,\nPloS one 11 (2016), no. 2, Article ID e0150171.","DOI":"10.1371\/journal.pone.0150171"},{"key":"2023033111373295703_j_cmam-2020-0130_ref_003","doi-asserted-by":"crossref","unstructured":"K. K. Chen, J. H. Tu and C. W. Rowley,\nVariants of dynamic mode decomposition: Boundary condition, Koopman, and Fourier analyses,\nJ. Nonlinear Sci. 22 (2012), no. 6, 887\u2013915.","DOI":"10.1007\/s00332-012-9130-9"},{"key":"2023033111373295703_j_cmam-2020-0130_ref_004","doi-asserted-by":"crossref","unstructured":"W. Chen, H. Sidky and A. L. Ferguson,\nNonlinear discovery of slow molecular modes using state-free reversible vampnets,\nJ. Chem. Phys. 150 (2019), no. 21, Article ID 214114.","DOI":"10.1063\/1.5092521"},{"key":"2023033111373295703_j_cmam-2020-0130_ref_005","doi-asserted-by":"crossref","unstructured":"J. D. Chodera and F. No\u00e9,\nMarkov state models of biomolecular conformational dynamics,\nCurr. Opin. Struct. Biol. 25 (2014), 135\u2013144.","DOI":"10.1016\/j.sbi.2014.04.002"},{"key":"2023033111373295703_j_cmam-2020-0130_ref_006","doi-asserted-by":"crossref","unstructured":"R. R. Coifman and S. Lafon,\nDiffusion maps,\nAppl. Comput. Harmon. Anal. 21 (2006), no. 1, 5\u201330.","DOI":"10.1016\/j.acha.2006.04.006"},{"key":"2023033111373295703_j_cmam-2020-0130_ref_007","doi-asserted-by":"crossref","unstructured":"N. D. Conrad, M. Weber and C. Sch\u00fctte,\nFinding dominant structures of nonreversible Markov processes,\nMultiscale Model. Simul. 14 (2016), no. 4, 1319\u20131340.","DOI":"10.1137\/15M1032272"},{"key":"2023033111373295703_j_cmam-2020-0130_ref_008","doi-asserted-by":"crossref","unstructured":"P. Deuflhard and M. Weber,\nRobust Perron cluster analysis in conformation dynamics,\nLinear Algebra Appl. 398 (2005), 161\u2013184.","DOI":"10.1016\/j.laa.2004.10.026"},{"key":"2023033111373295703_j_cmam-2020-0130_ref_009","doi-asserted-by":"crossref","unstructured":"P. Drineas and M. W. Mahoney,\nOn the Nystr\u00f6m method for approximating a Gram matrix for improved kernel-based learning,\nJ. Mach. Learn. Res. 6 (2005), 2153\u20132175.","DOI":"10.1007\/11503415_22"},{"key":"2023033111373295703_j_cmam-2020-0130_ref_010","doi-asserted-by":"crossref","unstructured":"K. Fackeldey, P. Koltai, P. N\u00e9vir, H. Rust, A. Schild and M. Weber,\nFrom metastable to coherent sets\u2014time-discretization schemes,\nChaos 29 (2019), no. 1, Article ID 012101.","DOI":"10.1063\/1.5058128"},{"key":"2023033111373295703_j_cmam-2020-0130_ref_011","unstructured":"K. Fukumizu, L. Song and A. Gretton,\nKernel Bayes\u2019 rule: Bayesian inference with positive definite kernels,\nJ. Mach. Learn. Res. 14 (2013), 3753\u20133783."},{"key":"2023033111373295703_j_cmam-2020-0130_ref_012","doi-asserted-by":"crossref","unstructured":"D. Goswami, E. Thackray and D. A. Paley,\nConstrained Ulam dynamic mode decomposition: Approximation of the Perron\u2013Frobenius operator for deterministic and stochastic systems,\nIEEE Control Syst. Lett. 2 (2018), no. 4, 809\u2013814.","DOI":"10.1109\/LCSYS.2018.2849552"},{"key":"2023033111373295703_j_cmam-2020-0130_ref_013","doi-asserted-by":"crossref","unstructured":"J. Han, L. Zhang and W. E,\nSolving many-electron Schr\u00f6dinger equation using deep neural networks,\nJ. Comput. Phys. 399 (2019), Article ID 108929.","DOI":"10.1016\/j.jcp.2019.108929"},{"key":"2023033111373295703_j_cmam-2020-0130_ref_014","unstructured":"Y. Hashimoto, I. Ishikawa, M. Ikeda, F. Komura and Y. Kawahara,\nKernel mean embeddings of von Neumann-algebra-valued measures, preprint (2020), https:\/\/arxiv.org\/abs\/2007.14698."},{"key":"2023033111373295703_j_cmam-2020-0130_ref_015","unstructured":"Y. Hashimoto, I. Ishikawa, M. Ikeda, Y. Matsuo and Y. Kawahara,\nKrylov subspace method for nonlinear dynamical systems with random noise,\nJ. Mach. Learn. Res. 21 (2020), Paper No. 172."},{"key":"2023033111373295703_j_cmam-2020-0130_ref_016","doi-asserted-by":"crossref","unstructured":"T. Hastie, R. Tibshirani and J. Friedman,\nThe Elements of Statistical Learning. Vol. 1,\nSpringer Ser. Statist.,\nSpringer, New York, 2001.","DOI":"10.1007\/978-0-387-21606-5_1"},{"key":"2023033111373295703_j_cmam-2020-0130_ref_017","doi-asserted-by":"crossref","unstructured":"J. Hermann, Z. Sch\u00e4tzle and F. No\u00e9,\nDeep-neural-network solution of the electronic Schr\u00f6dinger equation,\nNature Chem. 12 (2020), no. 10, 891\u2013897.","DOI":"10.1038\/s41557-020-0544-y"},{"key":"2023033111373295703_j_cmam-2020-0130_ref_018","doi-asserted-by":"crossref","unstructured":"I. T. Jolliffe and J. Cadima,\nPrincipal component analysis: A review and recent developments,\nPhilos. Trans. Roy. Soc. A 374 (2016), no. 2065, 10.1098\/rsta.2015.0202.","DOI":"10.1098\/rsta.2015.0202"},{"key":"2023033111373295703_j_cmam-2020-0130_ref_019","doi-asserted-by":"crossref","unstructured":"V. R. Joseph and L. Kang,\nRegression-based inverse distance weighting with applications to computer experiments,\nTechnometrics 53 (2011), no. 3, 254\u2013265.","DOI":"10.1198\/TECH.2011.09154"},{"key":"2023033111373295703_j_cmam-2020-0130_ref_020","doi-asserted-by":"crossref","unstructured":"O. Junge and P. Koltai,\nDiscretization of the Frobenius\u2013Perron operator using a sparse Haar tensor basis: The sparse Ulam method,\nSIAM J. Numer. Anal. 47 (2009), no. 5, 3464\u20133485.","DOI":"10.1137\/080716864"},{"key":"2023033111373295703_j_cmam-2020-0130_ref_021","unstructured":"Y. Kawahara,\nDynamic mode decomposition with reproducing kernels for koopman spectral analysis,\nAdvances in Neural Information Processing Systems,\nACM, New York (2016), 911\u2013919."},{"key":"2023033111373295703_j_cmam-2020-0130_ref_022","doi-asserted-by":"crossref","unstructured":"S. Klus, P. Koltai and C. Sch\u00fctte,\nOn the numerical approximation of the Perron\u2013Frobenius and Koopman operator,\nJ. Comput. Dyn. 3 (2016), no. 1, 51\u201379.","DOI":"10.3934\/jcd.2016003"},{"key":"2023033111373295703_j_cmam-2020-0130_ref_023","doi-asserted-by":"crossref","unstructured":"S. Klus, F. N\u00fcske and B. Hamzi,\nKernel-based approximation of the Koopman generator and Schr\u00f6dinger operator,\nEntropy 22 (2020), no. 7, Paper No. 722.","DOI":"10.3390\/e22070722"},{"key":"2023033111373295703_j_cmam-2020-0130_ref_024","doi-asserted-by":"crossref","unstructured":"S. Klus, F. N\u00fcske, P. Koltai, H. Wu, I. Kevrekidis, C. Sch\u00fctte and F. No\u00e9,\nData-driven model reduction and transfer operator approximation,\nJ. Nonlinear Sci. 28 (2018), no. 3, 985\u20131010.","DOI":"10.1007\/s00332-017-9437-7"},{"key":"2023033111373295703_j_cmam-2020-0130_ref_025","doi-asserted-by":"crossref","unstructured":"S. Klus, F. N\u00fcske, S. Peitz, J.-H. Niemann, C. Clementi and C. Sch\u00fctte,\nData-driven approximation of the Koopman generator: Model reduction, system identification, and control,\nPhys. D 406 (2020), Article ID 132416.","DOI":"10.1016\/j.physd.2020.132416"},{"key":"2023033111373295703_j_cmam-2020-0130_ref_026","doi-asserted-by":"crossref","unstructured":"S. Klus, I. Schuster and K. Muandet,\nEigendecompositions of transfer operators in reproducing kernel Hilbert spaces,\nJ. Nonlinear Sci. 30 (2020), no. 1, 283\u2013315.","DOI":"10.1007\/s00332-019-09574-z"},{"key":"2023033111373295703_j_cmam-2020-0130_ref_027","doi-asserted-by":"crossref","unstructured":"A. Konrad, B. Y. Zhao, A. D. Joseph and R. Ludwig,\nA Markov-based channel model algorithm for wireless networks,\nWireless Netw. 9 (2003), no. 3, 189\u2013199.","DOI":"10.1023\/A:1022869025953"},{"key":"2023033111373295703_j_cmam-2020-0130_ref_028","doi-asserted-by":"crossref","unstructured":"M. Korda and I. Mezi\u0107,\nOn convergence of extended dynamic mode decomposition to the Koopman operator,\nJ. Nonlinear Sci. 28 (2018), no. 2, 687\u2013710.","DOI":"10.1007\/s00332-017-9423-0"},{"key":"2023033111373295703_j_cmam-2020-0130_ref_029","doi-asserted-by":"crossref","unstructured":"J. N. Kutz, S. L. Brunton, B. W. Brunton and J. L. Proctor,\nDynamic Mode Decomposition: Data-Driven Modeling of Complex Systems,\nSociety for Industrial and Applied Mathematics (SIAM), Philadelphia, 2016.","DOI":"10.1137\/1.9781611974508"},{"key":"2023033111373295703_j_cmam-2020-0130_ref_030","doi-asserted-by":"crossref","unstructured":"S. Lafon, Y. Keller and R. R. Coifman,\nData fusion and multicue data matching by diffusion maps,\nIEEE Trans. Pattern Anal. Mach. Intell. 28 (2006), no. 11, 1784\u20131797.","DOI":"10.1109\/TPAMI.2006.223"},{"key":"2023033111373295703_j_cmam-2020-0130_ref_031","doi-asserted-by":"crossref","unstructured":"P. L. Lai and C. Fyfe,\nKernel and nonlinear canonical correlation analysis,\nInt. J. Neural Syst. 10 (2000), no. 05, 365\u2013377.","DOI":"10.1142\/S012906570000034X"},{"key":"2023033111373295703_j_cmam-2020-0130_ref_032","doi-asserted-by":"crossref","unstructured":"Q. Li, F. Dietrich, E. M. Bollt and I. G. Kevrekidis,\nExtended dynamic mode decomposition with dictionary learning: A data-driven adaptive spectral decomposition of the Koopman operator,\nChaos 27 (2017), no. 10, Article ID 103111.","DOI":"10.1063\/1.4993854"},{"key":"2023033111373295703_j_cmam-2020-0130_ref_033","doi-asserted-by":"crossref","unstructured":"O. Lindenbaum, M. Salhov, A. Yeredor and A. Averbuch,\nGaussian bandwidth selection for manifold learning and classification,\nData Min. Knowl. Discov. 34 (2020), no. 6, 1676\u20131712.","DOI":"10.1007\/s10618-020-00692-x"},{"key":"2023033111373295703_j_cmam-2020-0130_ref_034","doi-asserted-by":"crossref","unstructured":"O. Lindenbaum, A. Yeredor and A. Averbuch,\nBandwidth selection for kernel-based classification,\n2016 IEEE International Conference on the Science of Electrical Engineering (ICSEE),\nIEEE Press, Piscataway (2016), 1\u20135.","DOI":"10.1109\/ICSEE.2016.7806089"},{"key":"2023033111373295703_j_cmam-2020-0130_ref_035","doi-asserted-by":"crossref","unstructured":"B. Lusch, J. N. Kutz and S. L. Brunton,\nDeep learning for universal linear embeddings of nonlinear dynamics,\nNature Commun. 9 (2018), no. 1, 1\u201310.","DOI":"10.1038\/s41467-018-07210-0"},{"key":"2023033111373295703_j_cmam-2020-0130_ref_036","unstructured":"A. Mardt, L. Pasquali, F. No\u00e9 and H. Wu,\nDeep learning Markov and Koopman models with physical constraints,\nProc. Mach. Learn. Res. (PMLR) 107 (2020), 451\u2013475."},{"key":"2023033111373295703_j_cmam-2020-0130_ref_037","doi-asserted-by":"crossref","unstructured":"A. Mardt, L. Pasquali, H. Wu and F. No\u00e9,\nVampnets for deep learning of molecular kinetics,\nNature Commun. 9 (2018), no. 1, 1\u201311.","DOI":"10.1038\/s41467-017-02388-1"},{"key":"2023033111373295703_j_cmam-2020-0130_ref_038","doi-asserted-by":"crossref","unstructured":"R. T. McGibbon and V. S. Pande,\nVariational cross-validation of slow dynamical modes in molecular kinetics,\nJ. Chem. Phys. 142 (2015), no. 12, Article ID 124105.","DOI":"10.1063\/1.4916292"},{"key":"2023033111373295703_j_cmam-2020-0130_ref_039","doi-asserted-by":"crossref","unstructured":"I. Mezi\u0107,\nAnalysis of fluid flows via spectral properties of the Koopman operator,\nAnnu. Rev. Fluid Mech. 45 (2013), 357\u2013378.","DOI":"10.1146\/annurev-fluid-011212-140652"},{"key":"2023033111373295703_j_cmam-2020-0130_ref_040","doi-asserted-by":"crossref","unstructured":"K. Mills, M. Spanner and I. Tamblyn,\nDeep learning and the Schr\u00f6dinger equation,\nPhys. Rev. A 96 (2017), no. 4, Article ID 042113.","DOI":"10.1103\/PhysRevA.96.042113"},{"key":"2023033111373295703_j_cmam-2020-0130_ref_041","doi-asserted-by":"crossref","unstructured":"H. Q. Minh, P. Niyogi and Y. Yao,\nMercer\u2019s theorem, feature maps, and smoothing,\nLearning Theory,\nLecture Notes in Comput. Sci. 4005,\nSpringer, Berlin (2006), 154\u2013168.","DOI":"10.1007\/11776420_14"},{"key":"2023033111373295703_j_cmam-2020-0130_ref_042","doi-asserted-by":"crossref","unstructured":"L. Molgedey and H. G. Schuster,\nSeparation of a mixture of independent signals using time delayed correlations,\nPhys. Rev. Lett. 72 (1994), 3634\u20133637.","DOI":"10.1103\/PhysRevLett.72.3634"},{"key":"2023033111373295703_j_cmam-2020-0130_ref_043","doi-asserted-by":"crossref","unstructured":"J. R. Moreno, G. Carleo and A. Georges,\nDeep learning the Hohenberg-Kohn maps of density functional theory,\nPhys. Rev. Lett. 125 (2020), no. 7, Article ID 076402.","DOI":"10.1103\/PhysRevLett.125.076402"},{"key":"2023033111373295703_j_cmam-2020-0130_ref_044","doi-asserted-by":"crossref","unstructured":"R. Nagai, R. Akashi and O. Sugino,\nCompleting density functional theory by machine learning hidden messages from molecules,\nnpj Comput. Mat. 6 (2020), no. 1, 1\u20138.","DOI":"10.1038\/s41524-020-0310-0"},{"key":"2023033111373295703_j_cmam-2020-0130_ref_045","doi-asserted-by":"crossref","unstructured":"J. Nelson, R. Tiwari and S. Sanvito,\nMachine learning density functional theory for the Hubbard model,\nPhys. Rev. B 99 (2019), no. 7, Article ID 075132.","DOI":"10.1103\/PhysRevB.99.075132"},{"key":"2023033111373295703_j_cmam-2020-0130_ref_046","doi-asserted-by":"crossref","unstructured":"F. No\u00e9 and F. N\u00fcske,\nA variational approach to modeling slow processes in stochastic dynamical systems,\nMultiscale Model. Simul. 11 (2013), no. 2, 635\u2013655.","DOI":"10.1137\/110858616"},{"key":"2023033111373295703_j_cmam-2020-0130_ref_047","doi-asserted-by":"crossref","unstructured":"F. Nuske, B. G. Keller, G. Perezhernandez, A. S. J. S. Mey and F. Noe,\nVariational approach to molecular kinetics,\nJ. Chem. Theory Comput. 10 (2014), no. 4, 1739\u20131752.","DOI":"10.1021\/ct4009156"},{"key":"2023033111373295703_j_cmam-2020-0130_ref_048","doi-asserted-by":"crossref","unstructured":"F. N\u00faske, R. Schneider, F. Vitalini and F. No\u00e9,\nVariational tensor approach for approximating the rare event kinetics of macromolecular systems,\nJ. Chem. Phys. 144 (2016), no. 5, 149\u2013153.","DOI":"10.1063\/1.4940774"},{"key":"2023033111373295703_j_cmam-2020-0130_ref_049","doi-asserted-by":"crossref","unstructured":"G. Perezhernandez, F. Paul, T. Giorgino, G. De Fabritiis and F. Noe,\nIdentification of slow molecular order parameters for Markov model construction,\nJ. Chem. Phys. 139 (2013), Article ID 015102.","DOI":"10.1063\/1.4811489"},{"key":"2023033111373295703_j_cmam-2020-0130_ref_050","unstructured":"W. H. Press, S. A. Teukolsky, W. T. Vetterling and B. P. Flannery,\nNumerical Recipes: The Art of Scientific Computing,\nCambridge University, Cambridge, 2007."},{"key":"2023033111373295703_j_cmam-2020-0130_ref_051","doi-asserted-by":"crossref","unstructured":"J.-H. Prinz, H. Wu, M. Sarich, B. Keller, M. Senne, M. Held, J. D. Chodera, C. Sch\u00fctte and F. No\u00e9,\nMarkov models of molecular kinetics: Generation and validation,\nJ. Chem. Phys. 134 (2011), no. 17, Article ID 174105.","DOI":"10.1063\/1.3565032"},{"key":"2023033111373295703_j_cmam-2020-0130_ref_052","unstructured":"A. Rahimi and B. Recht,\nRandom features for large-scale kernel machines,\nAdvances in Neural Information Processing Systems,\nACM, New York (2008), 1177\u20131184."},{"key":"2023033111373295703_j_cmam-2020-0130_ref_053","doi-asserted-by":"crossref","unstructured":"S. R\u00f6blitz and M. Weber,\nFuzzy spectral clustering by PCCA+: Application to Markov state models and data classification,\nAdv. Data Anal. Classif. 7 (2013), no. 2, 147\u2013179.","DOI":"10.1007\/s11634-013-0134-6"},{"key":"2023033111373295703_j_cmam-2020-0130_ref_054","doi-asserted-by":"crossref","unstructured":"P. J. Schmid,\nDynamic mode decomposition of numerical and experimental data,\nJ. Fluid Mech. 656 (2010), 5\u201328.","DOI":"10.1017\/S0022112010001217"},{"key":"2023033111373295703_j_cmam-2020-0130_ref_055","doi-asserted-by":"crossref","unstructured":"C. Sch\u00fctte, A. Fischer, W. Huisinga and P. Deuflhard,\nA direct approach to conformational dynamics based on hybrid Monte Carlo,\nJ. Comput. Phys. 151 (1999), 146\u2013168.","DOI":"10.1006\/jcph.1999.6231"},{"key":"2023033111373295703_j_cmam-2020-0130_ref_056","doi-asserted-by":"crossref","unstructured":"C. R. Schwantes and V. S. Pande,\nImprovements in Markov state model construction reveal many non native interactions in the folding of ntl9,\nJ. Chem. Theory Comput. 9 (2013), no. 4, 2000\u20132009.","DOI":"10.1021\/ct300878a"},{"key":"2023033111373295703_j_cmam-2020-0130_ref_057","doi-asserted-by":"crossref","unstructured":"C. R. Schwantes and V. S. Pande,\nModeling molecular kinetics with tica and the kernel trick,\nJ. Chem. Theory Comput. 11 (2015), no. 2, 600\u2013608.","DOI":"10.1021\/ct5007357"},{"key":"2023033111373295703_j_cmam-2020-0130_ref_058","doi-asserted-by":"crossref","unstructured":"A. S. Sharma, I. Mezi and B. J. Mckeon,\nCorrespondence between koopman mode decomposition, resolvent mode decomposition, and invariant solutions of the Navier\u2013Stokes equations,\nPhys. Rev. Fluids 1 (2016), Article ID 03240.","DOI":"10.1103\/PhysRevFluids.1.032402"},{"key":"2023033111373295703_j_cmam-2020-0130_ref_059","doi-asserted-by":"crossref","unstructured":"A. Singer, R. Erban, I. G. Kevrekidis and R. R. Coifman,\nDetecting intrinsic slow variables in stochastic dynamical systems by anisotropic diffusion maps,\nProc. Natl. Acad. Sci. 106 (2009), no. 38, 16090\u201316095.","DOI":"10.1073\/pnas.0905547106"},{"key":"2023033111373295703_j_cmam-2020-0130_ref_060","doi-asserted-by":"crossref","unstructured":"A. Smola, A. Gretton, L. Song and B. Sch\u00f6lkopf,\nA Hilbert space embedding for distributions,\nInternational Conference on Algorithmic Learning Theory,\nSpringer, Berlin (2007), 13\u201331.","DOI":"10.1007\/978-3-540-75225-7_5"},{"key":"2023033111373295703_j_cmam-2020-0130_ref_061","doi-asserted-by":"crossref","unstructured":"L. Song, K. Fukumizu and A. Gretton,\nKernel embeddings of conditional distributions: A unified kernel framework for nonparametric inference in graphical models,\nIEEE Signal Process. Mag. 30 (2013), no. 4, 98\u2013111.","DOI":"10.1109\/MSP.2013.2252713"},{"key":"2023033111373295703_j_cmam-2020-0130_ref_062","doi-asserted-by":"crossref","unstructured":"L. Song, J. Huang, A. Smola and K. Fukumizu,\nHilbert space embeddings of conditional distributions with applications to dynamical systems,\nProceedings of the 26th Annual International Conference on Machine Learning,\nACM, New York (2009), 961\u2013968.","DOI":"10.1145\/1553374.1553497"},{"key":"2023033111373295703_j_cmam-2020-0130_ref_063","unstructured":"B. K. Sriperumbudur, K. Fukumizu and G. R. G. Lanckriet,\nUniversality, characteristic kernels and RKHS embedding of measures,\nJ. Mach. Learn. Res. 12 (2011), 2389\u20132410."},{"key":"2023033111373295703_j_cmam-2020-0130_ref_064","unstructured":"A. Szabo and N. S. Ostlund,\nModern Quantum Chemistry: Introduction to Advanced Electronic Structure Theory,\nCourier Corporation, North Chelmsford, 2012."},{"key":"2023033111373295703_j_cmam-2020-0130_ref_065","doi-asserted-by":"crossref","unstructured":"J. H. Tu, C. W. Rowley, D. M. Luchtenburg, S. L. Brunton and J. N. Kutz,\nOn dynamic mode decomposition: theory and applications,\nJ. Comput. Dyn. 1 (2014), no. 2, 391\u2013421.","DOI":"10.3934\/jcd.2014.1.391"},{"key":"2023033111373295703_j_cmam-2020-0130_ref_066","doi-asserted-by":"crossref","unstructured":"M. O. Williams, I. G. Kevrekidis and C. W. Rowley,\nA data-driven approximation of the Koopman operator: Extending dynamic mode decomposition,\nJ. Nonlinear Sci. 25 (2015), no. 6, 1307\u20131346.","DOI":"10.1007\/s00332-015-9258-5"},{"key":"2023033111373295703_j_cmam-2020-0130_ref_067","doi-asserted-by":"crossref","unstructured":"M. O. Williams, C. W. Rowley and I. G. Kevrekidis,\nA kernel-based method for data-driven Koopman spectral analysis,\nJ. Comput. Dyn. 2 (2015), no. 2, 247\u2013265.","DOI":"10.3934\/jcd.2015005"},{"key":"2023033111373295703_j_cmam-2020-0130_ref_068","unstructured":"H. Wu, A. Mardt, L. Pasquali and F. Noe,\nDeep generative Markov state models,\nAdvances in Neural Information Processing Systems,\nACM. New York (2018), 3975\u20133984."},{"key":"2023033111373295703_j_cmam-2020-0130_ref_069","doi-asserted-by":"crossref","unstructured":"H. Wu and F. No\u00e9,\nGaussian Markov transition models of molecular kinetics,\nJ. Chem. Phys. 142 (2015), no. 8, Article ID 084104.","DOI":"10.1063\/1.4913214"},{"key":"2023033111373295703_j_cmam-2020-0130_ref_070","doi-asserted-by":"crossref","unstructured":"H. Wu and F. No\u00e9,\nVariational approach for learning Markov processes from time series data,\nJ. Nonlinear Sci. 30 (2020), no. 1, 23\u201366.","DOI":"10.1007\/s00332-019-09567-y"},{"key":"2023033111373295703_j_cmam-2020-0130_ref_071","doi-asserted-by":"crossref","unstructured":"H. Wu, F. N\u00faske, F. Paul, S. Klus, P. Koltai and F. No\u00e9,\nVariational Koopman models: Slow collective variables and molecular kinetics from short off-equilibrium simulations,\nJ. Chem. Phys. 146 (2017), no. 15, Article ID 154104.","DOI":"10.1063\/1.4979344"},{"key":"2023033111373295703_j_cmam-2020-0130_ref_072","doi-asserted-by":"crossref","unstructured":"H. Wu, F. Paul, C. Wehmeyer and F. No\u00e9,\nMultiensemble Markov models of molecular thermodynamics and kinetics,\nProc. Natl. Acad. Sci. 113 (2016), no. 23, E3221\u2013E3230.","DOI":"10.1073\/pnas.1525092113"},{"key":"2023033111373295703_j_cmam-2020-0130_ref_073","doi-asserted-by":"crossref","unstructured":"M. Yue, J. J. Han and K. S. Trivedi,\nComposite performance and availability analysis of wirelesscommunication networks,\nIEEE Trans. Vehicular Technol. 50 (2001), no. 5, 1216\u20131223.","DOI":"10.1109\/25.950322"}],"container-title":["Computational Methods in Applied Mathematics"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.degruyter.com\/document\/doi\/10.1515\/cmam-2020-0130\/xml","content-type":"application\/xml","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/www.degruyter.com\/document\/doi\/10.1515\/cmam-2020-0130\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2023,3,31]],"date-time":"2023-03-31T15:18:19Z","timestamp":1680275899000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.degruyter.com\/document\/doi\/10.1515\/cmam-2020-0130\/html"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2021,6,9]]},"references-count":73,"journal-issue":{"issue":"3","published-online":{"date-parts":[[2021,5,6]]},"published-print":{"date-parts":[[2021,7,1]]}},"alternative-id":["10.1515\/cmam-2020-0130"],"URL":"https:\/\/doi.org\/10.1515\/cmam-2020-0130","relation":{},"ISSN":["1609-4840","1609-9389"],"issn-type":[{"type":"print","value":"1609-4840"},{"type":"electronic","value":"1609-9389"}],"subject":[],"published":{"date-parts":[[2021,6,9]]}}}