Measuring the Probabilistic Photometric Redshifts of X-ray Quasars Based on the Quantile Regression of Ensembles of Decision Trees

Meshcheryakov, A. V.; Glazkova, V. V.; Gerasimov, S. V.; Mashechkin, I. V.

doi:10.1134/s1063773718120058

Cited by 21 publications

(8 citation statements)

References 44 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It means that we could improve our result, if, in addition to good photometry for the entire sample, we could increase the training sample of bright objects by the time when eROSITA survey will be available. Photo-z computed via ML for X-ray selected sources in 3XMM-DR6 and 3XMM-DR7 where recently presented also in Ruiz et al (2018) and Meshcheryakov et al (2018), respectively. While we are in overall agreement with the first, our results are less optimistic than those obtained by the second group.…”

Section: X-ray Depthmentioning

confidence: 99%

“…This makes ML a natural choice for the computation of photo-z for the very large forthcoming deep and wide surveys such as Euclid (Laureijs 2010) and LSST (Ivezić et al 2019), in which the computation of photo-z will be a real challenge. For AGN, the next challenge is presented by the ∼ 3 million sources that eROSITA (extended Roentgen Survey with an Imaging Telescope Array; Merloni et al 2012), the primary instrument on the Russian Spektrum-Roentgen-Gamma (SRG) mission, will detect. eROSITA will provide an all-sky X-ray survey every 6 months for 4 years, with a final expected depth of 1 × 10 −14 erg/cm 2 /s (3 × 10 −15 erg/cm 2 /s at the poles) which is about 30 times deeper than ROSAT (Voges et al 1999;Boller et al 2016) in the soft band (0.5-2 keV).…”

mentioning

confidence: 99%

See 1 more Smart Citation

Photometric redshifts for X-ray-selected active galactic nuclei in the eROSITA era

Brescia

Salvato

Cavuoti

et al. 2019

Monthly Notices of the Royal Astronomical Society

View full text Add to dashboard Cite

With the launch of eROSITA, successfully occurred on July 13th, 2019, we are facing the challenge of computing reliable photometric redshifts for 3 million of AGNs over the entire sky, having available only patchy and inhomogeneous ancillary data. While we have a good understanding of the photo-z quality obtainable for AGN using SEDfitting technique, we tested the capability of Machine Learning (ML), usually reliable in computing photo-z for QSO in wide and shallow areas with rich spectroscopic samples. Using MLPQNA as example of ML, we computed photo-z for the X-ray selected sources in Stripe 82X, using the publicly available photometric and spectroscopic catalogues. Stripe 82X is at least as deep as eROSITA will be and wide enough to include also rare and bright AGNs. In addition, the availability of ancillary data mimics what can be available in the whole sky. We found that when optical, NIR and MIR data are available, ML and SED-fitting perform comparably well in terms of overall accuracy, realistic redshift probability density functions and fraction of outliers, although they are not the same for the two methods. The results could further improve if the photometry available is accurate and including morphological information. Assuming that we can gather sufficient spectroscopy to build a representative training sample, with the current photometry coverage we can obtain reliable photo-z for a large fraction of sources in the Southern Hemisphere well before the spectroscopic follow-up, thus timely enabling the eROSITA science return. The photo-z catalogue is released here.

show abstract

Section: X-ray Depthmentioning

confidence: 99%

mentioning

confidence: 99%

Photometric redshifts for X-ray-selected active galactic nuclei in the eROSITA era

Brescia

Salvato

Cavuoti

et al. 2019

Monthly Notices of the Royal Astronomical Society

View full text Add to dashboard Cite

show abstract

“…• supervised feed-forward neural networks (Collister and Lahav, 2004;Vanzella et al, 2004;Brescia et al, 2013Brescia et al, , 2014Brescia et al, , 2015Brescia et al, , 2019Cavuoti et al, 2014;Almosallam et al, 2016;Sadeh et al, 2016); • self-adaptive methods for the detection and removal of anomalies from photometric and spectroscopic data (Hoyle et al, 2015;Baron and Poznanski, 2017;Reis et al, 2019); • Support Vector Machines (Zheng and Zhang, 2012;Zhang and Zhao, 2014;Han et al, 2016;Jones and Singal, 2017); • tree-based (Carrasco Kind and Brunner, 2013;Jouvel et al, 2017;Meshcheryakov et al, 2018); • k-Nearest Neighbors (kNN) (Graham et al, 2018;Curran, 2020); • Gaussian processes (Bonfield et al, 2010;Almosallam et al, 2016); • Mixture Density Networks (Ansari et al, 2020); • unsupervised models for clustering and for estimating the coverage of the parameter space (Way and Klose, 2012;Masters et al, 2015;Stensbo-Smidt et al, 2017) or for calibration purposes (Hildebrandt et al, 2010;Masters et al, 2015;Wright et al, 2020); • deep Neural Networks, especially relevant for the photo-z prediction from images Chong and Yang, 2019;Pasquet et al, 2019); • hybrid methods for the selection of photometric redshifts considered particularly accurate and useful for cosmological purposes (Bonnett et al, 2016;Leistedt and Hogg, 2017;…”

Section: General Aspects Of the Photo-z Estimation With MLmentioning

confidence: 99%

Photometric Redshifts With Machine Learning, Lights and Shadows on a Complex Data Science Use Case

Brescia

Cavuoti

Razim

et al. 2021

Front. Astron. Space Sci.

View full text Add to dashboard Cite

The importance of the current role of data-driven science is constantly increasing within Astrophysics, due to the huge amount of multi-wavelength data collected every day, characterized by complex and high-volume information requiring efficient and, as much as possible, automated exploration tools. Furthermore, to accomplish main and legacy science objectives of future or incoming large and deep survey projects, such as James Webb Space Telescope (JWST), James Webb Space Telescope (LSST), and Euclid, a crucial role is played by an accurate estimation of photometric redshifts, whose knowledge would permit the detection and analysis of extended and peculiar sources by disentangling low-z from high-z sources and would contribute to solve the modern cosmological discrepancies. The recent photometric redshift data challenges, organized within several survey projects, like LSST and Euclid, pushed the exploitation of the observed multi-wavelength and multi-dimensional data or ad hoc simulated data to improve and optimize the photometric redshifts prediction and statistical characterization based on both Spectral Energy Distribution (SED) template fitting and machine learning methodologies. They also provided a new impetus in the investigation of hybrid and deep learning techniques, aimed at conjugating the positive peculiarities of different methodologies, thus optimizing the estimation accuracy and maximizing the photometric range coverage, which are particularly important in the high-z regime, where the spectroscopic ground truth is poorly available. In such a context, we summarize what was learned and proposed in more than a decade of research.

show abstract

“…The resulting list of optical candidates was processed by the SRGz system, which operates over the entire Eastern Galactic hemisphere region of the eROSITA X-ray survey and automatically analyzes photometry and positions of optical objects in the Xray sources vicinity. SRGz use tree-based machine learning algorithms (Gradient Boosting and Random Forest, see Meshcheryakov et al, 2018), which are trained on samples of quasars, galaxies, and stars from the SDSS spectroscopic catalog, sample of distant z > 5 quasars (Ross & Cross, 2020) and sample of GAIA DR2 stars associated with 3XMM DR8 sources. For more details on the principles of operation of SRGz and the algorithms implemented in it see Meshcheryakov (2021).…”

Section: Modelmentioning

confidence: 99%

Discovery of the most X-ray luminous quasar SRGE J170245.3+130104 at redshift z$\approx5.5$

Khorunzhev,

Meshcheryakov,

Medvedev

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

SRGE J170245.3+130104 was discovered by the eROSITA telescope aboard the SRG space observatory on March 13-15, 2020 during the first half-year scan of its all-sky X-ray survey. The optical counterpart of the X-ray source was photometrically identified as a distant quasar candidate at z ≈ 5.5. Follow-up spectroscopic observations, done in August/September 2020 with the SCORPIO-II instrument at the BTA 6-m telescope, confirmed that SRGE J170245.3+130104 is a distant quasar at redshift zspec = 5.466 ± 0.003. The X-ray luminosity of the quasar during the first half-year scan of the eROSITA all-sky survey was 3.6 +2.1 −1.5 × 10 46 erg/s (in the 2-10 keV energy range), whereas its X-ray spectrum could be described by a power law with a slope of Γ = 1.8 +0.9 −0.8 . Six months later (September 13-14, 2020), during the second half-year scan of the eROSITA all-sky survey, the quasar was detected again and its X-ray luminosity had decreased by a factor of 2 (at the ≈ 1.9σ confidence level). SRGE J170245.3+130104 proves to be the most X-ray luminous among all known X-ray quasars at z > 5. It is also one of the radio-loudest distant quasars (with radio-loudness R ∼ 10 3 ), which may imply that it is a blazar. In the Appendix, we present the list of all z > 5 quasars detected in X-rays to date.

show abstract

Measuring the Probabilistic Photometric Redshifts of X-ray Quasars Based on the Quantile Regression of Ensembles of Decision Trees

Cited by 21 publications

References 44 publications

Photometric redshifts for X-ray-selected active galactic nuclei in the eROSITA era

Photometric redshifts for X-ray-selected active galactic nuclei in the eROSITA era

Photometric Redshifts With Machine Learning, Lights and Shadows on a Complex Data Science Use Case

Discovery of the most X-ray luminous quasar SRGE J170245.3+130104 at redshift z$\approx5.5$

Contact Info

Product

Resources

About