Machine learning methods for constructing probabilistic<i>Fermi</i>-LAT catalogs

Bhat, Aakash; Malyshev, Dmitry

doi:10.1051/0004-6361/202140766

Cited by 10 publications

(7 citation statements)

References 60 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…2. Overall, our classifier yields a relatively large fraction of sources with P PSR γ > 0.5, namely around 45%, compared to other works (e.g., Bhat & Malyshev 2022;Coronado-Blázquez 2022). This is likely caused by our approach of using a balanced classifier, which implicitly assumes that both classes are a priori equally likely, as well by our attempt to not bias ourselves against faint sources (see Appendix A).…”

Section: Classification Of γ-Ray Sourcesmentioning

confidence: 85%

“…It has been shown numerous times (Salvetti 2014;Saz Parkinson et al 2016;Kaur et al 2019;Kerby et al 2021a,b;Germani et al 2021;Bhat & Malyshev 2022;Coronado-Blázquez 2022), that the properties of γ-ray sources in the Fermi-LAT source catalogs can be used to predict their source type, and in particular to differentiate between pulsar-like and blazar-like sources. Typically, the primary differences between the two classes were found to be a stronger spectral downward curvature for pulsars, and stronger variability for blazars.…”

Section: Classification Of γ-Ray Sourcesmentioning

confidence: 99%

“…The type of classifier we used for this task was a random forest classifier, as was frequently done before (Salvetti 2014;Saz Parkinson et al 2016;Bhat & Malyshev 2022). In order to tackle the strongly imbalanced nature of the training set Fig.…”

Section: Classification Of γ-Ray Sourcesmentioning

confidence: 99%

See 2 more Smart Citations

Searching for X-ray counterparts of unassociated Fermi-LAT sources and rotation-powered pulsars with SRG/eROSITA

Mayer,

Becker

2024

A&A

View full text Add to dashboard Cite

The latest source catalog of the Fermi -LAT telescope contains more than 7000 gamma -ray sources at giga-electronvolt energies, with the two dominant source classes thought to be blazars and rotation-powered pulsars. Despite continuous follow-up efforts, around 2600 sources have no known multiwavelength association. Our target is the identification of possible (young and recycled) pulsar candidates in the sample of unassociated gamma -ray sources via their characteristic X-ray and gamma -ray emission. To achieve this, we cross-matched the Fermi -LAT catalog with the catalog of X-ray sources in the western Galactic hemisphere from the first four all-sky surveys of eROSITA on the Spektrum-Roentgen-Gamma ( SRG ) mission. We complement this by identifying X-ray counterparts of known pulsars detected at gamma -ray and radio energies in the eROSITA data. We used a Bayesian cross-matching scheme to construct a probabilistic catalog of possible pulsar-type X-ray counterparts to Fermi -LAT sources. Our method combines the overlap of X-ray and gamma -ray source positions with a probabilistic classification (into pulsar and blazar candidates) of each source based on its gamma -ray properties and a prediction on the X-ray flux of pulsar- or blazar-type counterparts. Finally, an optical and infrared counterpart search was performed to exclude coronally emitting stars and active galactic nuclei from our catalog. We provide a catalog of our prior gamma -ray-based classifications of all 2600 unassociated sources in the Fermi -LAT catalog, with around equal numbers of pulsar and blazar candidates. Our final list of candidate X-ray counterparts to suspected new high-energy pulsars, cleaned for spurious detections and sources with obvious non-pulsar counterparts, contains around 900 X-ray sources, the vast majority of which lie in the 95<!PCT!> gamma -ray error ellipse. We predict between 30 and 40 new pulsars among our top 200 candidates, with around equal numbers of young and recycled pulsars. This candidate list may serve as input to future follow-up campaigns, looking directly for pulsations or for the orbital modulation of possible binary companions, where it may allow for a drastic reduction in the number of candidate locations to search. We furthermore detect the X-ray counterparts of 15 known rotation-powered pulsars, which were not seen in X-rays before.

show abstract

Section: Classification Of γ-Ray Sourcesmentioning

confidence: 85%

Section: Classification Of γ-Ray Sourcesmentioning

confidence: 99%

See 1 more Smart Citation

Searching for X-ray counterparts of unassociated Fermi-LAT sources and rotation-powered pulsars with SRG/eROSITA

Mayer,

Becker

2024

A&A

View full text Add to dashboard Cite

show abstract

“…Another straightforward use of our results could be a guided source search and identification program via multi-wavelength studies, which would also help transforming some of these candidate source directions into bona fide sources. In turn, these studies may further benefit of machine learning techniques that have been recently proposed to ease the probabilistic classification of unassociated sources [31].…”

Section: Jcap03(2024)055mentioning

confidence: 99%

Deepening gamma-ray point-source catalogues with sub-threshold information

Amerio,

Calore,

Serpico

et al. 2024

J. Cosmol. Astropart. Phys.

View full text Add to dashboard Cite

We propose a novel statistical method to extend Fermi-LAT catalogues of high-latitude γ-ray sources below their nominal threshold. To do so, we rely on the determination of the differential source-count distribution of sub-threshold sources which only provides the statistical flux distribution of faint sources. By simulating ensembles of synthetic skies, we assess quantitatively the likelihood for pixels in the sky with relatively low-test statistics to be due to sources, therefore complementing the source-count distribution with spatial information. Besides being useful to orient efforts towards multi-messenger and multi-wavelength identification of new γ-ray sources, we expect the results to be especially advantageous for statistical applications such as cross-correlation analyses.

show abstract

“…Machine learning classifiers have been used in recent years to study UNID sources [28][29][30][31][32][33][34][35] and to search for potential dark subhalos. Using XGBoost classifiers on sources in the third Fermi-LAT catalog, ref.…”

Section: Jcap07(2023)033mentioning

confidence: 99%

Searching for dark matter subhalos in the Fermi-LAT catalog with Bayesian neural networks

Butter

Krämer

Manconi

et al. 2023

J. Cosmol. Astropart. Phys.

View full text Add to dashboard Cite

About a third of the γ-ray sources detected by the Fermi Large Area Telescope (Fermi-LAT) remain unidentified, and some of these could be exotic objects such as dark matter subhalos. We present a search for these sources using Bayesian neural network classification methods applied to the latest 4FGL-DR3 Fermi-LAT catalog. We first simulate the γ-ray properties of dark matter subhalos using models from N-body simulations and semi-analytical approaches to the subhalo distribution. We then assess the detectability of this sample in the 4FGL-DR3 catalog using the Fermi-LAT analysis tools. We train our Bayesian neural network to identify candidate dark matter subhalos among the unidentified sources in the 4FGL-DR3 catalog. Our results allow us to derive conservative bounds on the dark matter annihilation cross section by excluding unidentified sources classified as astrophysical-like by our networks. We estimate the number of candidate dark matter subhalos for different dark matter masses and provide a publicly available list for further investigation. Our bounds on the dark matter annihilation cross section are comparable to previous results and become particularly competitive at high dark matter masses.

show abstract

Machine learning methods for constructing probabilisticFermi-LAT catalogs

Cited by 10 publications

References 60 publications

Searching for X-ray counterparts of unassociated Fermi-LAT sources and rotation-powered pulsars with SRG/eROSITA

Searching for X-ray counterparts of unassociated Fermi-LAT sources and rotation-powered pulsars with SRG/eROSITA

Deepening gamma-ray point-source catalogues with sub-threshold information

Searching for dark matter subhalos in the Fermi-LAT catalog with Bayesian neural networks

Contact Info

Product

Resources

About