{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2025,2,21]],"date-time":"2025-02-21T23:58:44Z","timestamp":1740182324112,"version":"3.37.3"},"reference-count":34,"publisher":"IOP Publishing","issue":"2","license":[{"start":{"date-parts":[[2023,6,20]],"date-time":"2023-06-20T00:00:00Z","timestamp":1687219200000},"content-version":"vor","delay-in-days":19,"URL":"http:\/\/creativecommons.org\/licenses\/by\/4.0"},{"start":{"date-parts":[[2023,6,20]],"date-time":"2023-06-20T00:00:00Z","timestamp":1687219200000},"content-version":"tdm","delay-in-days":19,"URL":"https:\/\/iopscience.iop.org\/info\/page\/text-and-data-mining"}],"funder":[{"name":"National Science Foundation","award":["2112085"],"award-info":[{"award-number":["2112085"]}]}],"content-domain":{"domain":["iopscience.iop.org"],"crossmark-restriction":false},"short-container-title":["Mach. Learn.: Sci. Technol."],"published-print":{"date-parts":[[2023,6,1]]},"abstract":"<jats:title>Abstract<\/jats:title>\n               <jats:p>In many scenarios, it is necessary to monitor a complex system via a time-series of observations and determine when anomalous exogenous events have occurred so that relevant actions can be taken. Determining whether current observations are abnormal is challenging. It requires learning an extrapolative probabilistic model of the dynamics from historical data, and using a limited number of current observations to make a classification. We leverage recent advances in long-term probabilistic forecasting, namely <jats:italic>Deep Probabilistic Koopman<\/jats:italic>, to build a general method for classifying anomalies in multi-dimensional time-series data. We also show how to utilize models with domain knowledge of the dynamics to reduce type I and type II error. We demonstrate our proposed method on the important real-world task of global atmospheric pollution monitoring, integrating it with NASA\u2019s Global Earth Observing System Model. The system successfully detects localized anomalies in air quality due to events such as COVID-19 lockdowns and wildfires.<\/jats:p>","DOI":"10.1088\/2632-2153\/acdd50","type":"journal-article","created":{"date-parts":[[2023,6,9]],"date-time":"2023-06-09T22:40:59Z","timestamp":1686350459000},"page":"025033","update-policy":"https:\/\/doi.org\/10.1088\/crossmark-policy","source":"Crossref","is-referenced-by-count":1,"title":["Koopman-inspired approach for identification of exogenous anomalies in nonstationary time-series data"],"prefix":"10.1088","volume":"4","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-5120-1350","authenticated-orcid":true,"given":"Alex","family":"Mallen","sequence":"first","affiliation":[]},{"given":"Christoph A","family":"Keller","sequence":"additional","affiliation":[]},{"ORCID":"https:\/\/orcid.org\/0000-0002-6004-2275","authenticated-orcid":true,"given":"J Nathan","family":"Kutz","sequence":"additional","affiliation":[]}],"member":"266","published-online":{"date-parts":[[2023,6,20]]},"reference":[{"volume":"vol 3","year":"2000","author":"Shumway","key":"mlstacdd50bib1"},{"key":"mlstacdd50bib2","doi-asserted-by":"publisher","first-page":"1","DOI":"10.5555\/3546258.3546299","article-title":"From fourier to koopman: spectral methods for long-term time series prediction","volume":"22","author":"Lange","year":"2021","journal-title":"J. Mach. Learn. Res."},{"article-title":"Deep probabilistic koopman: long-term time-series forecasting under periodic uncertainties","year":"2021","author":"Mallen","key":"mlstacdd50bib3"},{"key":"mlstacdd50bib4","doi-asserted-by":"publisher","DOI":"10.1088\/2632-2153\/ac3de0","article-title":"Deeptime: a python library for machine learning dynamical models from time series data","volume":"3","author":"Hoffmann","year":"2021","journal-title":"Mach. Learn.: Sci. Technol."},{"year":"2016","author":"Kutz","key":"mlstacdd50bib5"},{"year":"2019","author":"Brunton","key":"mlstacdd50bib6"},{"key":"mlstacdd50bib7","first-page":"pp 15","article-title":"Machine learning for sequential data: a review","author":"Dietterich","year":"2002"},{"key":"mlstacdd50bib8","doi-asserted-by":"publisher","first-page":"76","DOI":"10.1111\/joes.12429","article-title":"Machine learning advances for time series forecasting","volume":"37","author":"Masini","year":"2021","journal-title":"J. Econ. Surv."},{"year":"2013","author":"Kutz","key":"mlstacdd50bib9"},{"year":"1999","author":"Mallat","key":"mlstacdd50bib10"},{"key":"mlstacdd50bib11","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":"J. Fluid Mech."},{"key":"mlstacdd50bib12","doi-asserted-by":"publisher","first-page":"115","DOI":"10.1017\/S0022112009992059","article-title":"Spectral analysis of nonlinear flows","volume":"645","author":"Rowley","year":"2009","journal-title":"J. Fluid Mech."},{"key":"mlstacdd50bib13","doi-asserted-by":"publisher","first-page":"380","DOI":"10.1137\/M1124176","article-title":"Variable projection methods for an optimized dynamic mode decomposition","volume":"17","author":"Askham","year":"2018","journal-title":"SIAM J. Appl. Dyn. Syst."},{"key":"mlstacdd50bib14","doi-asserted-by":"publisher","first-page":"315","DOI":"10.1073\/pnas.17.5.315","article-title":"Hamiltonian systems and transformation in Hilbert space","volume":"17","author":"Koopman","year":"1931","journal-title":"Proc. Natl Acad. Sci. USA"},{"key":"mlstacdd50bib15","doi-asserted-by":"publisher","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 Dyn."},{"article-title":"Modern koopman theory for dynamical systems","year":"2021","author":"Brunton","key":"mlstacdd50bib16"},{"key":"mlstacdd50bib17","doi-asserted-by":"publisher","DOI":"10.1063\/1.4772195","article-title":"Applied koopmanism","volume":"22","author":"Budi\u0161i\u0107","year":"2012","journal-title":"Chaos"},{"key":"mlstacdd50bib18","doi-asserted-by":"publisher","first-page":"357","DOI":"10.1146\/annurev-fluid-011212-140652","article-title":"Analysis of fluid flows via spectral properties of the koopman operator","volume":"45","author":"Mezi\u0107","year":"2013","journal-title":"Annu. Rev. Fluid Mech."},{"key":"mlstacdd50bib19","doi-asserted-by":"publisher","first-page":"2096","DOI":"10.1137\/17M1125236","article-title":"Ergodic theory, dynamic mode decomposition and computation of spectral properties of the koopman operator","volume":"16","author":"Arbabi","year":"2017","journal-title":"SIAM J. Appl. Dyn. Syst."},{"key":"mlstacdd50bib20","doi-asserted-by":"publisher","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":"Brunton","year":"2016","journal-title":"Proc. Natl Acad. Sci."},{"key":"mlstacdd50bib21","doi-asserted-by":"publisher","first-page":"22445","DOI":"10.1073\/pnas.1906995116","article-title":"Data-driven discovery of coordinates and governing equations","volume":"116","author":"Champion","year":"2019","journal-title":"Proc. Natl Acad. Sci."},{"key":"mlstacdd50bib22","doi-asserted-by":"publisher","first-page":"251","DOI":"10.1016\/0893-6080(91)90009-T","article-title":"Approximation capabilities of multilayer feedforward networks","volume":"4","author":"Hornik","year":"1991","journal-title":"Neural Netw."},{"year":"2016","author":"Goodfellow","key":"mlstacdd50bib23"},{"key":"mlstacdd50bib24","doi-asserted-by":"publisher","first-page":"1235","DOI":"10.1162\/neco_a_01199","article-title":"A review of recurrent neural networks: Lstm cells and network architectures","volume":"31","author":"Yong","year":"2019","journal-title":"Neural Comput."},{"key":"mlstacdd50bib25","doi-asserted-by":"publisher","first-page":"1735","DOI":"10.1162\/neco.1997.9.8.1735","article-title":"Long short-term memory","volume":"9","author":"Hochreiter","year":"1997","journal-title":"Neural Comput."},{"key":"mlstacdd50bib26","doi-asserted-by":"publisher","DOI":"10.1029\/2020MS002413","article-title":"Description of the nasa geos composition forecast modeling system geos-cf v1.0","volume":"13","author":"Keller","year":"2021","journal-title":"J. Adv. Model. Earth Syst."},{"year":"2017","author":"Brasseur","key":"mlstacdd50bib27"},{"article-title":"Modern Koopman theory for dynamical systems","year":"2021","author":"Brunton","key":"mlstacdd50bib28"},{"key":"mlstacdd50bib29","doi-asserted-by":"publisher","first-page":"255","DOI":"10.1073\/pnas.18.3.255","article-title":"Dynamical systems of continuous spectra","volume":"18","author":"Koopman","year":"1932","journal-title":"Proc. Natl Acad. Sci. USA"},{"key":"mlstacdd50bib30","doi-asserted-by":"publisher","first-page":"539","DOI":"10.1146\/annurev.fl.25.010193.002543","article-title":"The proper orthogonal decomposition in the analysis of turbulent flows","volume":"25","author":"Berkooz","year":"1993","journal-title":"Annu. Rev. Fluid Mech."},{"key":"mlstacdd50bib31","doi-asserted-by":"publisher","first-page":"5040","DOI":"10.1109\/TIT.2014.2323359","article-title":"The optimal hard threshold for singular values is 43","volume":"60","author":"Gavish","year":"2014","journal-title":"IEEE Trans. Inf. Theory"},{"key":"mlstacdd50bib32","doi-asserted-by":"publisher","first-page":"289","DOI":"10.1007\/s13253-019-00355-5","article-title":"Hypothesis tests for principal component analysis when variables are standardized","volume":"24","author":"Forkman","year":"2019","journal-title":"J. Agric. Biol. Environ. Stat."},{"year":"2021","author":"Keller","key":"mlstacdd50bib33"},{"key":"mlstacdd50bib34","doi-asserted-by":"publisher","first-page":"3555","DOI":"10.5194\/acp-21-3555-2021","article-title":"Global impact of covid-19 restrictions on the surface concentrations of nitrogen dioxide and ozone","volume":"21","author":"Keller","year":"2021","journal-title":"Atmos. Chem. Phys."}],"container-title":["Machine Learning: Science and Technology"],"original-title":[],"link":[{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/acdd50","content-type":"text\/html","content-version":"am","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/acdd50\/pdf","content-type":"application\/pdf","content-version":"am","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/acdd50","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/acdd50\/pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/acdd50\/pdf","content-type":"application\/pdf","content-version":"am","intended-application":"syndication"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/acdd50\/pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/acdd50\/pdf","content-type":"application\/pdf","content-version":"am","intended-application":"similarity-checking"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/acdd50\/pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2023,6,20]],"date-time":"2023-06-20T08:02:48Z","timestamp":1687248168000},"score":1,"resource":{"primary":{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/acdd50"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2023,6,1]]},"references-count":34,"journal-issue":{"issue":"2","published-online":{"date-parts":[[2023,6,20]]},"published-print":{"date-parts":[[2023,6,1]]}},"URL":"https:\/\/doi.org\/10.1088\/2632-2153\/acdd50","relation":{},"ISSN":["2632-2153"],"issn-type":[{"type":"electronic","value":"2632-2153"}],"subject":[],"published":{"date-parts":[[2023,6,1]]},"assertion":[{"value":"Koopman-inspired approach for identification of exogenous anomalies in nonstationary time-series data","name":"article_title","label":"Article Title"},{"value":"Machine Learning: Science and Technology","name":"journal_title","label":"Journal Title"},{"value":"paper","name":"article_type","label":"Article Type"},{"value":"\u00a9 2023 The Author(s). Published by IOP Publishing Ltd","name":"copyright_information","label":"Copyright Information"},{"value":"2022-11-27","name":"date_received","label":"Date Received","group":{"name":"publication_dates","label":"Publication dates"}},{"value":"2023-06-09","name":"date_accepted","label":"Date Accepted","group":{"name":"publication_dates","label":"Publication dates"}},{"value":"2023-06-20","name":"date_epub","label":"Online publication date","group":{"name":"publication_dates","label":"Publication dates"}}]}}