{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,3,30]],"date-time":"2026-03-30T16:25:52Z","timestamp":1774887952443,"version":"3.50.1"},"reference-count":55,"publisher":"IOP Publishing","issue":"4","license":[{"start":{"date-parts":[[2022,12,28]],"date-time":"2022-12-28T00:00:00Z","timestamp":1672185600000},"content-version":"vor","delay-in-days":27,"URL":"http:\/\/creativecommons.org\/licenses\/by\/4.0"},{"start":{"date-parts":[[2022,12,28]],"date-time":"2022-12-28T00:00:00Z","timestamp":1672185600000},"content-version":"tdm","delay-in-days":27,"URL":"https:\/\/iopscience.iop.org\/info\/page\/text-and-data-mining"}],"funder":[{"name":"URA Visiting Scholars Program"},{"DOI":"10.13039\/100006208","name":"High Energy Physics","doi-asserted-by":"crossref","id":[{"id":"10.13039\/100006208","id-type":"DOI","asserted-by":"crossref"}]},{"DOI":"10.13039\/100000015","name":"U.S. Department of Energy","doi-asserted-by":"crossref","id":[{"id":"10.13039\/100000015","id-type":"DOI","asserted-by":"crossref"}]},{"name":"Fermi Research Alliance, LLC","award":["DE-AC02-07CH11359"],"award-info":[{"award-number":["DE-AC02-07CH11359"]}]}],"content-domain":{"domain":["iopscience.iop.org"],"crossmark-restriction":false},"short-container-title":["Mach. Learn.: Sci. Technol."],"published-print":{"date-parts":[[2022,12,1]]},"abstract":"<jats:title>Abstract<\/jats:title>\n               <jats:p>With the advent of billion-galaxy surveys with complex data, the need of the hour is to efficiently model galaxy spectral energy distributions (SEDs) with robust uncertainty quantification. The combination of simulation-based inference (SBI) and amortized neural posterior estimation (NPE) has been successfully used to analyse simulated and real galaxy photometry both precisely and efficiently. In this work, we utilise this combination and build on existing literature to analyse simulated noisy galaxy spectra. Here, we demonstrate a proof-of-concept study of spectra that is (a) an efficient analysis of galaxy SEDs and inference of galaxy parameters with physically interpretable uncertainties; and (b) amortized calculations of posterior distributions of said galaxy parameters at the modest cost of a few galaxy fits with Markov chain Monte Carlo (MCMC) methods. We utilise the SED generator and inference framework Prospector to generate simulated spectra, and train a dataset of 2 \u00d7 10<jats:sup>6<\/jats:sup> spectra (corresponding to a five-parameter SED model) with NPE. We show that SBI\u2014with its combination of fast and amortized posterior estimations\u2014is capable of inferring accurate galaxy stellar masses and metallicities. Our uncertainty constraints are comparable to or moderately weaker than traditional inverse-modelling with Bayesian MCMC methods (e.g. 0.17 and 0.26 dex in stellar mass and metallicity for a given galaxy, respectively). We also find that our inference framework conducts rapid SED inference (0.9\u20131.2 \u00d7 10<jats:sup>5<\/jats:sup> galaxy spectra via SBI\/NPE at the cost of 1 MCMC-based fit). With this work, we set the stage for further work that focuses of SED fitting of galaxy spectra with SBI, in the era of JWST galaxy survey programs and the wide-field Roman Space Telescope spectroscopic surveys.<\/jats:p>","DOI":"10.1088\/2632-2153\/ac98f4","type":"journal-article","created":{"date-parts":[[2022,10,10]],"date-time":"2022-10-10T22:20:25Z","timestamp":1665440425000},"page":"04LT04","update-policy":"https:\/\/doi.org\/10.1088\/crossmark-policy","source":"Crossref","is-referenced-by-count":16,"title":["DIGS: deep inference of galaxy spectra with neural posterior estimation"],"prefix":"10.1088","volume":"3","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-3475-7648","authenticated-orcid":true,"given":"Gourav","family":"Khullar","sequence":"first","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-6706-8972","authenticated-orcid":false,"given":"Brian","family":"Nord","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-1281-7192","authenticated-orcid":true,"given":"Aleksandra","family":"\u0106iprijanovi\u0107","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-5040-093X","authenticated-orcid":false,"given":"Jason","family":"Poh","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Fei","family":"Xu","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"266","published-online":{"date-parts":[[2022,12,28]]},"reference":[{"key":"mlstac98f4bib1","doi-asserted-by":"publisher","first-page":"18","DOI":"10.3847\/1538-4365\/aae9f0","article-title":"The dark energy survey: data release 1","volume":"239","author":"Abbott","year":"2018","journal-title":"Astrophys. J. Suppl. Ser."},{"key":"mlstac98f4bib2","article-title":"The DESI experiment part I: science,targeting, and survey design","author":"Aghamousa","year":"2016"},{"key":"mlstac98f4bib3","doi-asserted-by":"publisher","first-page":"3","DOI":"10.3847\/1538-4365\/ab929e","article-title":"The 16th data release of the sloan digital sky surveys: first release from the APOGEE-2 southern survey and full release of eBOSS spectra","volume":"249","author":"Ahumada","year":"2020","journal-title":"Astrophys. J. Suppl. Ser."},{"key":"mlstac98f4bib4","doi-asserted-by":"publisher","first-page":"4440","DOI":"10.1093\/mnras\/stz1960","article-title":"Fast likelihood-free cosmology with neural density estimators and active learning","volume":"488","author":"Alsing","year":"2019","journal-title":"Mon. Not. R. Astron. Soc."},{"key":"mlstac98f4bib5","doi-asserted-by":"publisher","first-page":"57","DOI":"10.1088\/0004-637X\/770\/1\/57","article-title":"The average star formation histories of galaxies in dark matter halos from z = 0-8","volume":"770","author":"Behroozi","year":"2013","journal-title":"Astrophys. J."},{"key":"mlstac98f4bib6","doi-asserted-by":"publisher","first-page":"17","DOI":"10.3847\/1538-4357\/ab07af","article-title":"MOSFIRE spectroscopy of quiescent galaxies at 1.5 < z < 2.5. II. Star formation histories and galaxy quenching","volume":"874","author":"Belli","year":"2019","journal-title":"Astrophys. J."},{"key":"mlstac98f4bib7","doi-asserted-by":"publisher","first-page":"44","DOI":"10.3847\/1538-4357\/ab04a2","article-title":"How to measure galaxy star formation histories. I. Parametric models","volume":"873","author":"Carnall","year":"2019","journal-title":"Astrophys. J."},{"key":"mlstac98f4bib8","doi-asserted-by":"publisher","first-page":"102","DOI":"10.3847\/0004-637X\/823\/2\/102","article-title":"Mesa isochrones and stellar tracks (MIST). I. Solar-scaled models","volume":"823","author":"Choi","year":"2016","journal-title":"Astrophys. J."},{"key":"mlstac98f4bib9","doi-asserted-by":"publisher","first-page":"677","DOI":"10.1093\/mnras\/stab1677","article-title":"DeepMerge\u2014II. Building robust deep learning algorithms for merging galaxy identification across domains","volume":"506","author":"\u0106iprijanovi\u0107","year":"2021","journal-title":"Mon. Not. R. Astron. Soc."},{"key":"mlstac98f4bib10","doi-asserted-by":"publisher","DOI":"10.1088\/2632-2153\/ac7f1a","article-title":"DeepAdversaries: examining the robustness of deep learning models for galaxy morphology classification","volume":"3","author":"\u0106iprijanovi\u0107","year":"2022","journal-title":"Mach. Learn.: Sci. Technol."},{"key":"mlstac98f4bib11","doi-asserted-by":"publisher","first-page":"833","DOI":"10.1088\/0004-637X\/712\/2\/833","article-title":"The propagation of uncertainties in stellar population synthesis modeling. III. Model calibration, comparison and evaluation","volume":"712","author":"Conroy","year":"2010","journal-title":"Astrophys. J."},{"key":"mlstac98f4bib12","article-title":"The frontier of simulation-based inference","author":"Cranmer","year":"2019"},{"key":"mlstac98f4bib13","first-page":"pp 1","article-title":"A comprehensive survey on domain adaptation for visual applications","author":"Csurka","year":"2017"},{"key":"mlstac98f4bib14","doi-asserted-by":"publisher","first-page":"168","DOI":"10.3847\/1538-3881\/ab089d","article-title":"Overview of the DESI legacy imaging surveys","volume":"157","author":"Dey","year":"2019","journal-title":"Astron. J."},{"key":"mlstac98f4bib15","doi-asserted-by":"publisher","first-page":"A95","DOI":"10.1051\/0004-6361\/201116842","article-title":"An updated MILES stellar library and stellar population models","volume":"532","author":"Falc\u00f3n-Barroso","year":"2011","journal-title":"Astron. Astrophys."},{"key":"mlstac98f4bib16","doi-asserted-by":"publisher","first-page":"81","DOI":"10.1086\/307513","article-title":"The age and metallicity range of early-type galaxies in clusters","volume":"521","author":"Ferreras","year":"1999","journal-title":"Astrophys. J."},{"key":"mlstac98f4bib17","doi-asserted-by":"publisher","first-page":"306","DOI":"10.1086\/670067","article-title":"emcee: the MCMC hammer","volume":"125","author":"Foreman-Mackey","year":"2013","journal-title":"Publ. Astron. Soc. Pac."},{"key":"mlstac98f4bib18","article-title":"Ensemble deep learning: a review","author":"Ganaie","year":"2021"},{"key":"mlstac98f4bib19","doi-asserted-by":"publisher","first-page":"16","DOI":"10.1086\/679742","article-title":"On the recovery of galaxy properties from SED fitting solutions","volume":"127","author":"Gladis Magris","year":"2015","journal-title":"Publ. Astron. Soc. Pac."},{"key":"mlstac98f4bib20","article-title":"Automatic posterior transformation for likelihood-free inference","author":"Greenberg","year":"2019"},{"key":"mlstac98f4bib21","article-title":"The prime focus spectrograph galaxy evolution survey","author":"Greene","year":"2022"},{"key":"mlstac98f4bib22","doi-asserted-by":"crossref","DOI":"10.3847\/1538-4357\/ac7b84","article-title":"Accelerated Bayesian SED modeling using amortized neural posterior estimation","author":"Hahn","year":"2022"},{"key":"mlstac98f4bib23","doi-asserted-by":"publisher","first-page":"19","DOI":"10.1088\/0067-0049\/208\/2\/19","article-title":"Nine-year Wilkinson microwave anisotropy probe (WMAP) observations: cosmological parameter results","volume":"208","author":"Hinshaw","year":"2013","journal-title":"Astrophys. J. Suppl. Ser."},{"key":"mlstac98f4bib24","doi-asserted-by":"publisher","first-page":"5689","DOI":"10.1093\/mnras\/stab3437","article-title":"Accurate x-ray timing in the presence of systematic biases with simulation-based inference","volume":"511","author":"Huppenkothen","year":"2022","journal-title":"Mon. Not. R. Astron. Soc."},{"key":"mlstac98f4bib25","doi-asserted-by":"publisher","first-page":"111","DOI":"10.3847\/1538-4357\/ab042c","article-title":"LSST: from science drivers to reference design and anticipated data products","volume":"873","author":"Ivezi\u0107","year":"2019","journal-title":"Astrophys. J."},{"key":"mlstac98f4bib26","doi-asserted-by":"publisher","first-page":"22","DOI":"10.3847\/1538-4365\/abef67","article-title":"Stellar population inference with prospector","volume":"254","author":"Johnson","year":"2021","journal-title":"Astrophys. J. Suppl. Ser."},{"key":"mlstac98f4bib27","doi-asserted-by":"publisher","first-page":"042","DOI":"10.1088\/1475-7516\/2018\/02\/042","article-title":"Accelerating approximate Bayesian computation with quantile regression: application to cosmological redshift distributions","volume":"2018","author":"Kacprzak","year":"2018","journal-title":"J. Cosmol. Astropart. Phys."},{"key":"mlstac98f4bib28","article-title":"Characterizing stellar mass assembly and physical properties in the brightest galaxy in the redshift > 5 Universe","author":"Khullar","year":"2021"},{"key":"mlstac98f4bib29","doi-asserted-by":"publisher","first-page":"177","DOI":"10.3847\/1538-4357\/ac7c0c","article-title":"Synthesizing stellar populations in South Pole Telescope galaxy clusters. I. Ages of quiescent member galaxies at 0.3 < z < 1.4","volume":"934","author":"Khullar","year":"2022","journal-title":"Astrophys. J."},{"key":"mlstac98f4bib30","doi-asserted-by":"publisher","first-page":"231","DOI":"10.1046\/j.1365-8711.2001.04022.x","article-title":"On the variation of the initial mass function","volume":"322","author":"Kroupa","year":"2001","journal-title":"Mon. Not. R. Astron. Soc."},{"key":"mlstac98f4bib31","article-title":"A very early onset of massive galaxy formation","author":"Labbe","year":"2022"},{"key":"mlstac98f4bib32","article-title":"Early results from GLASS-JWST. X: rest-frame UV-optical properties of galaxies at 7 < z < 9","author":"Leethochawalit","year":"2022"},{"key":"mlstac98f4bib33","doi-asserted-by":"publisher","first-page":"140","DOI":"10.3847\/1538-4357\/ab1d5a","article-title":"An older, more quiescent universe from panchromatic SED fitting of the 3D-HST survey","volume":"877","author":"Leja","year":"2019b","journal-title":"Astrophys. J."},{"key":"mlstac98f4bib34","doi-asserted-by":"publisher","first-page":"3","DOI":"10.3847\/1538-4357\/ab133c","article-title":"How to measure galaxy star formation histories. II. Nonparametric models","volume":"876","author":"Leja","year":"2019a","journal-title":"Astrophys. J."},{"key":"mlstac98f4bib35","doi-asserted-by":"publisher","first-page":"170","DOI":"10.3847\/1538-4357\/aa5ffe","article-title":"Deriving physical properties from broadband photometry with prospector: description of the model and a demonstration of its accuracy using 129 galaxies in the local universe","volume":"837","author":"Leja","year":"2017a","journal-title":"Astrophys. J."},{"key":"mlstac98f4bib36","doi-asserted-by":"publisher","first-page":"3255","DOI":"10.1093\/mnras\/sty3217","article-title":"Deep learning of multi-element abundances from high-resolution spectroscopic data","volume":"483","author":"Leung","year":"2019","journal-title":"Mon. Not. R. Astron. Soc."},{"key":"mlstac98f4bib37","doi-asserted-by":"publisher","first-page":"5503","DOI":"10.1093\/mnras\/stz2851","article-title":"Learning the relationship between galaxies spectra and their star formation histories using convolutional neural networks and cosmological simulations","volume":"490","author":"Lovell","year":"2019","journal-title":"Mon. Not. R. Astron. Soc."},{"key":"mlstac98f4bib38","doi-asserted-by":"publisher","first-page":"1694","DOI":"10.1093\/mnras\/sty1024","article-title":"Simulating galaxies in the reionization era with FIRE-2: galaxy scaling relations, stellar mass functions and luminosity functions","volume":"478","author":"Ma","year":"2018","journal-title":"Mon. Not. R. Astron. Soc."},{"key":"mlstac98f4bib39","article-title":"Early results from GLASS-JWST XVI: discovering a bluer z \u223c 4-7 universe through UV slopes","author":"Nanayakkara","year":"2022"},{"key":"mlstac98f4bib40","article-title":"Fast \u03f5-free inference of simulation models with bayesian conditional density estimation","author":"Papamakarios","year":"2016"},{"key":"mlstac98f4bib41","article-title":"Masked autoregressive flow for density estimation","author":"Papamakarios","year":"2017"},{"key":"mlstac98f4bib42","doi-asserted-by":"crossref","DOI":"10.3847\/2041-8213\/ac8e06","article-title":"Rest-frame near-infrared sizes of galaxies at cosmic noon: objects in JWST\u2019s mirror are smaller than they appeared","author":"Suess","year":"2022a"},{"key":"mlstac98f4bib43","doi-asserted-by":"crossref","DOI":"10.3847\/1538-4357\/ac82b0","article-title":"Recovering the star formation histories of recently-quenched galaxies: the impact of model and prior choices","author":"Suess","year":"2022b"},{"key":"mlstac98f4bib44","doi-asserted-by":"publisher","first-page":"145","DOI":"10.1002\/cpa.21423","article-title":"A family of nonparametric density estimation algorithms","volume":"66","author":"Tabak","year":"2013","journal-title":"Commun. Pure Appl. Math."},{"key":"mlstac98f4bib45","doi-asserted-by":"publisher","first-page":"134","DOI":"10.3847\/1538-4357\/ac449b","article-title":"Fast, slow, early, late: quenching massive galaxies at z \u223c 0.8","volume":"926","author":"Tacchella","year":"2022","journal-title":"Astrophys. J."},{"key":"mlstac98f4bib46","article-title":"Validating Bayesian inference algorithms with simulation-based calibration","author":"Talts","year":"2018"},{"key":"mlstac98f4bib47","doi-asserted-by":"publisher","first-page":"2505","DOI":"10.21105\/joss.02505","article-title":"SBI: a toolkit for simulation-based inference","volume":"5","author":"Tejero-Cantero","year":"2020","journal-title":"J. Open Source Softw."},{"key":"mlstac98f4bib48","article-title":"Neural autoregressive distribution estimation","author":"Uria","year":"2016"},{"key":"mlstac98f4bib49","article-title":"Hands-on Bayesian neural networks\u2014a tutorial for deep learning users","author":"Valentin Jospin","year":"2020"},{"key":"mlstac98f4bib50","doi-asserted-by":"publisher","first-page":"135","DOI":"10.1016\/j.neucom.2018.05.083","article-title":"Deep visual domain adaptation: a survey","volume":"312","author":"Wang","year":"2018","journal-title":"Neurocomputing"},{"key":"mlstac98f4bib51","doi-asserted-by":"publisher","first-page":"1","DOI":"10.3847\/1538-4357\/ac4973","article-title":"The high latitude spectroscopic survey on the Nancy Grace Roman Space Telescope","volume":"928","author":"Wang","year":"2022","journal-title":"Astrophys. J."},{"key":"mlstac98f4bib52","doi-asserted-by":"publisher","first-page":"107","DOI":"10.1086\/192096","article-title":"Comprehensive stellar population models and the disentanglement of age and metallicity effects","volume":"95","author":"Worthey","year":"1994","journal-title":"Astrophys. J. Suppl. Ser."},{"key":"mlstac98f4bib53","doi-asserted-by":"publisher","first-page":"78","DOI":"10.3847\/1538-4357\/836\/1\/78","article-title":"The spectral evolution of the first galaxies. III. Simulated James Webb Space Telescope spectra of reionization-epoch galaxies with Lyman-continuum leakage","volume":"836","author":"Zackrisson","year":"2017","journal-title":"Astrophys. J."},{"key":"mlstac98f4bib54","doi-asserted-by":"publisher","first-page":"262","DOI":"10.3847\/1538-3881\/abf42e","article-title":"Real-time likelihood-free inference of roman binary microlensing events with amortized neural posterior estimation","volume":"161","author":"Zhang","year":"2021","journal-title":"Astron. J."},{"key":"mlstac98f4bib55","doi-asserted-by":"publisher","first-page":"151","DOI":"10.3847\/1538-4357\/ac457d","article-title":"Simulation-based inference of reionization parameters from 3D tomographic 21 cm light-cone images","volume":"926","author":"Zhao","year":"2022","journal-title":"Astrophys. J."}],"container-title":["Machine Learning: Science and Technology"],"original-title":[],"link":[{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/ac98f4","content-type":"text\/html","content-version":"am","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/ac98f4\/pdf","content-type":"application\/pdf","content-version":"am","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/ac98f4","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/ac98f4\/pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/ac98f4\/pdf","content-type":"application\/pdf","content-version":"am","intended-application":"syndication"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/ac98f4\/pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/ac98f4\/pdf","content-type":"application\/pdf","content-version":"am","intended-application":"similarity-checking"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/ac98f4\/pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2022,12,28]],"date-time":"2022-12-28T09:58:52Z","timestamp":1672221532000},"score":1,"resource":{"primary":{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/ac98f4"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2022,12,1]]},"references-count":55,"journal-issue":{"issue":"4","published-online":{"date-parts":[[2022,12,28]]},"published-print":{"date-parts":[[2022,12,1]]}},"URL":"https:\/\/doi.org\/10.1088\/2632-2153\/ac98f4","relation":{},"ISSN":["2632-2153"],"issn-type":[{"value":"2632-2153","type":"electronic"}],"subject":[],"published":{"date-parts":[[2022,12,1]]},"assertion":[{"value":"DIGS: deep inference of galaxy spectra with neural posterior estimation","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 2022 The Author(s). Published by IOP Publishing Ltd","name":"copyright_information","label":"Copyright Information"},{"value":"2022-08-03","name":"date_received","label":"Date Received","group":{"name":"publication_dates","label":"Publication dates"}},{"value":"2022-10-10","name":"date_accepted","label":"Date Accepted","group":{"name":"publication_dates","label":"Publication dates"}},{"value":"2022-12-28","name":"date_epub","label":"Online publication date","group":{"name":"publication_dates","label":"Publication dates"}}]}}