HAYATE: Photometric redshift estimation by hybridising machine learning with template fitting
Machine learning photo-z methods, trained directly on spectroscopic redshifts, provide a viable alternative to traditional template fitting methods but may not generalise well on new data that deviates from that in the training set. In this work, we present a Hybrid Algorithm for WI(Y)de-range photo...
Gespeichert in:
| Hauptverfasser: | , , , , |
|---|---|
| Format: | Artikel |
| Sprache: | eng |
| Schlagworte: | |
| Online-Zugang: | Volltext bestellen |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| container_end_page | |
|---|---|
| container_issue | |
| container_start_page | |
| container_title | |
| container_volume | |
| creator | Tanigawa, Shingo Glazebrook, Karl Jacobs, Colin Labbe, Ivo Qin, Alex K |
| description | Machine learning photo-z methods, trained directly on spectroscopic
redshifts, provide a viable alternative to traditional template fitting methods
but may not generalise well on new data that deviates from that in the training
set. In this work, we present a Hybrid Algorithm for WI(Y)de-range photo-z
estimation with Artificial neural networks and TEmplate fitting (HAYATE), a
novel photo-z method that combines template fitting and data-driven approaches
and whose training loss is optimised in terms of both redshift point estimates
and probability distributions. We produce artificial training data from
low-redshift galaxy SEDs at z |
| doi_str_mv | 10.48550/arxiv.2402.00323 |
| format | Article |
| fullrecord | <record><control><sourceid>arxiv_GOX</sourceid><recordid>TN_cdi_arxiv_primary_2402_00323</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>2402_00323</sourcerecordid><originalsourceid>FETCH-LOGICAL-a673-a7b9602f6508128e8626671c4a7d765bcf3692b1334198601fe5e05f4c50a0c13</originalsourceid><addsrcrecordid>eNotj71OwzAURr0woMIDMOEXSLi2Yydhi6pCkSrBkIUBRY5zTa6Un8qxgLw9bWH6pDN8OoexOwFpVmgNDzb80FcqM5ApgJLqmn3sq_eq3j3yt36O84gxkOMBu6UnHzkukUYbaZ54u_J-bQN1tND0yUfrepqQD2jDdAbfFHsecTwONiL3FOOJ3rArb4cFb_93w-qnXb3dJ4fX55dtdUisyVVi87Y0IL3RUAhZYGGkMblwmc273OjWeWVK2QqlMlEWBoRHjaB95jRYcEJt2P3f7aWvOYaTdFibc2dz6VS_sWVNpw</addsrcrecordid><sourcetype>Open Access Repository</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype></control><display><type>article</type><title>HAYATE: Photometric redshift estimation by hybridising machine learning with template fitting</title><source>arXiv.org</source><creator>Tanigawa, Shingo ; Glazebrook, Karl ; Jacobs, Colin ; Labbe, Ivo ; Qin, Alex K</creator><creatorcontrib>Tanigawa, Shingo ; Glazebrook, Karl ; Jacobs, Colin ; Labbe, Ivo ; Qin, Alex K</creatorcontrib><description>Machine learning photo-z methods, trained directly on spectroscopic
redshifts, provide a viable alternative to traditional template fitting methods
but may not generalise well on new data that deviates from that in the training
set. In this work, we present a Hybrid Algorithm for WI(Y)de-range photo-z
estimation with Artificial neural networks and TEmplate fitting (HAYATE), a
novel photo-z method that combines template fitting and data-driven approaches
and whose training loss is optimised in terms of both redshift point estimates
and probability distributions. We produce artificial training data from
low-redshift galaxy SEDs at z<1.3, artificially redshifted up to z=5. We test
the model on data from the ZFOURGE surveys, demonstrating that HAYATE can
function as a reliable emulator of EAZY for the broad redshift range beyond the
region of sufficient spectroscopic completeness. The network achieves precise
photo-z estimations with smaller errors ($\sigma_{NMAD}$) than EAZY in the
initial low-z region (z<1.3), while being comparable even in the high-z
extrapolated regime (1.3<z<5). Meanwhile, it provides more robust photo-z
estimations than EAZY with the lower outlier rate ($\eta_{0.2}\lesssim 1\%$)
but runs $\sim100$ times faster than the original template fitting method. We
also demonstrate HAYATE offers more reliable redshift PDFs, showing a flatter
distribution of Probability Integral Transform scores than EAZY. The
performance is further improved using transfer learning with spec-z samples. We
expect that future large surveys will benefit from our novel methodology
applicable to observations over a wide redshift range.</description><identifier>DOI: 10.48550/arxiv.2402.00323</identifier><language>eng</language><subject>Physics - Cosmology and Nongalactic Astrophysics ; Physics - Instrumentation and Methods for Astrophysics</subject><creationdate>2024-01</creationdate><rights>http://creativecommons.org/licenses/by/4.0</rights><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><link.rule.ids>228,230,776,881</link.rule.ids><linktorsrc>$$Uhttps://arxiv.org/abs/2402.00323$$EView_record_in_Cornell_University$$FView_record_in_$$GCornell_University$$Hfree_for_read</linktorsrc><backlink>$$Uhttps://doi.org/10.48550/arXiv.2402.00323$$DView paper in arXiv$$Hfree_for_read</backlink></links><search><creatorcontrib>Tanigawa, Shingo</creatorcontrib><creatorcontrib>Glazebrook, Karl</creatorcontrib><creatorcontrib>Jacobs, Colin</creatorcontrib><creatorcontrib>Labbe, Ivo</creatorcontrib><creatorcontrib>Qin, Alex K</creatorcontrib><title>HAYATE: Photometric redshift estimation by hybridising machine learning with template fitting</title><description>Machine learning photo-z methods, trained directly on spectroscopic
redshifts, provide a viable alternative to traditional template fitting methods
but may not generalise well on new data that deviates from that in the training
set. In this work, we present a Hybrid Algorithm for WI(Y)de-range photo-z
estimation with Artificial neural networks and TEmplate fitting (HAYATE), a
novel photo-z method that combines template fitting and data-driven approaches
and whose training loss is optimised in terms of both redshift point estimates
and probability distributions. We produce artificial training data from
low-redshift galaxy SEDs at z<1.3, artificially redshifted up to z=5. We test
the model on data from the ZFOURGE surveys, demonstrating that HAYATE can
function as a reliable emulator of EAZY for the broad redshift range beyond the
region of sufficient spectroscopic completeness. The network achieves precise
photo-z estimations with smaller errors ($\sigma_{NMAD}$) than EAZY in the
initial low-z region (z<1.3), while being comparable even in the high-z
extrapolated regime (1.3<z<5). Meanwhile, it provides more robust photo-z
estimations than EAZY with the lower outlier rate ($\eta_{0.2}\lesssim 1\%$)
but runs $\sim100$ times faster than the original template fitting method. We
also demonstrate HAYATE offers more reliable redshift PDFs, showing a flatter
distribution of Probability Integral Transform scores than EAZY. The
performance is further improved using transfer learning with spec-z samples. We
expect that future large surveys will benefit from our novel methodology
applicable to observations over a wide redshift range.</description><subject>Physics - Cosmology and Nongalactic Astrophysics</subject><subject>Physics - Instrumentation and Methods for Astrophysics</subject><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2024</creationdate><recordtype>article</recordtype><sourceid>GOX</sourceid><recordid>eNotj71OwzAURr0woMIDMOEXSLi2Yydhi6pCkSrBkIUBRY5zTa6Un8qxgLw9bWH6pDN8OoexOwFpVmgNDzb80FcqM5ApgJLqmn3sq_eq3j3yt36O84gxkOMBu6UnHzkukUYbaZ54u_J-bQN1tND0yUfrepqQD2jDdAbfFHsecTwONiL3FOOJ3rArb4cFb_93w-qnXb3dJ4fX55dtdUisyVVi87Y0IL3RUAhZYGGkMblwmc273OjWeWVK2QqlMlEWBoRHjaB95jRYcEJt2P3f7aWvOYaTdFibc2dz6VS_sWVNpw</recordid><startdate>20240131</startdate><enddate>20240131</enddate><creator>Tanigawa, Shingo</creator><creator>Glazebrook, Karl</creator><creator>Jacobs, Colin</creator><creator>Labbe, Ivo</creator><creator>Qin, Alex K</creator><scope>GOX</scope></search><sort><creationdate>20240131</creationdate><title>HAYATE: Photometric redshift estimation by hybridising machine learning with template fitting</title><author>Tanigawa, Shingo ; Glazebrook, Karl ; Jacobs, Colin ; Labbe, Ivo ; Qin, Alex K</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-a673-a7b9602f6508128e8626671c4a7d765bcf3692b1334198601fe5e05f4c50a0c13</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2024</creationdate><topic>Physics - Cosmology and Nongalactic Astrophysics</topic><topic>Physics - Instrumentation and Methods for Astrophysics</topic><toplevel>online_resources</toplevel><creatorcontrib>Tanigawa, Shingo</creatorcontrib><creatorcontrib>Glazebrook, Karl</creatorcontrib><creatorcontrib>Jacobs, Colin</creatorcontrib><creatorcontrib>Labbe, Ivo</creatorcontrib><creatorcontrib>Qin, Alex K</creatorcontrib><collection>arXiv.org</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Tanigawa, Shingo</au><au>Glazebrook, Karl</au><au>Jacobs, Colin</au><au>Labbe, Ivo</au><au>Qin, Alex K</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>HAYATE: Photometric redshift estimation by hybridising machine learning with template fitting</atitle><date>2024-01-31</date><risdate>2024</risdate><abstract>Machine learning photo-z methods, trained directly on spectroscopic
redshifts, provide a viable alternative to traditional template fitting methods
but may not generalise well on new data that deviates from that in the training
set. In this work, we present a Hybrid Algorithm for WI(Y)de-range photo-z
estimation with Artificial neural networks and TEmplate fitting (HAYATE), a
novel photo-z method that combines template fitting and data-driven approaches
and whose training loss is optimised in terms of both redshift point estimates
and probability distributions. We produce artificial training data from
low-redshift galaxy SEDs at z<1.3, artificially redshifted up to z=5. We test
the model on data from the ZFOURGE surveys, demonstrating that HAYATE can
function as a reliable emulator of EAZY for the broad redshift range beyond the
region of sufficient spectroscopic completeness. The network achieves precise
photo-z estimations with smaller errors ($\sigma_{NMAD}$) than EAZY in the
initial low-z region (z<1.3), while being comparable even in the high-z
extrapolated regime (1.3<z<5). Meanwhile, it provides more robust photo-z
estimations than EAZY with the lower outlier rate ($\eta_{0.2}\lesssim 1\%$)
but runs $\sim100$ times faster than the original template fitting method. We
also demonstrate HAYATE offers more reliable redshift PDFs, showing a flatter
distribution of Probability Integral Transform scores than EAZY. The
performance is further improved using transfer learning with spec-z samples. We
expect that future large surveys will benefit from our novel methodology
applicable to observations over a wide redshift range.</abstract><doi>10.48550/arxiv.2402.00323</doi><oa>free_for_read</oa></addata></record> |
| fulltext | fulltext_linktorsrc |
| identifier | DOI: 10.48550/arxiv.2402.00323 |
| ispartof | |
| issn | |
| language | eng |
| recordid | cdi_arxiv_primary_2402_00323 |
| source | arXiv.org |
| subjects | Physics - Cosmology and Nongalactic Astrophysics Physics - Instrumentation and Methods for Astrophysics |
| title | HAYATE: Photometric redshift estimation by hybridising machine learning with template fitting |
| url | https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-02-06T22%3A48%3A42IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-arxiv_GOX&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=HAYATE:%20Photometric%20redshift%20estimation%20by%20hybridising%20machine%20learning%20with%20template%20fitting&rft.au=Tanigawa,%20Shingo&rft.date=2024-01-31&rft_id=info:doi/10.48550/arxiv.2402.00323&rft_dat=%3Carxiv_GOX%3E2402_00323%3C/arxiv_GOX%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_id=info:pmid/&rfr_iscdi=true |