scholarly journals ML-MOC: Machine Learning (kNN and GMM) based Membership determination for Open Clusters

2021 ◽  
Vol 502 (2) ◽  
pp. 2582-2599
Author(s):  
Manan Agarwal ◽  
Khushboo K Rao ◽  
Kaushar Vaidya ◽  
Souradeep Bhattacharya
Keyword(s):  
Machine Learning ◽  
Open Cluster ◽  
A Priori ◽  
Gaussian Mixture ◽  
Open Clusters ◽  
Proper Motions ◽  
Cluster Membership ◽  
Priori Information ◽  
Gaia Dr2 ◽  
Good Agreement

ABSTRACT The existing open-cluster membership determination algorithms are either prior dependent on some known parameters of clusters or are not automatable to large samples of clusters. In this paper, we present ml-moc, a new machine-learning-based approach to identify likely members of open clusters using the Gaia DR2 data and no a priori information about cluster parameters. We use the k-nearest neighbour (kNN) algorithm and the Gaussian mixture model (GMM) on high-precision proper motions and parallax measurements from the Gaia DR2 data to determine the membership probabilities of individual sources down to G ∼ 20 mag. To validate the developed method, we apply it to 15 open clusters: M67, NGC 2099, NGC 2141, NGC 2243, NGC 2539, NGC 6253, NGC 6405, NGC 6791, NGC 7044, NGC 7142, NGC 752, Blanco 1, Berkeley 18, IC 4651, and Hyades. These clusters differ in terms of their ages, distances, metallicities, and extinctions and cover a wide parameter space in proper motions and parallaxes with respect to the field population. The extracted members produce clean colour–magnitude diagrams and our astrometric parameters of the clusters are in good agreement with the values derived in previous work. The estimated degree of contamination in the extracted members ranges between 2 ${{\ \rm per\ cent}}$ and 12 ${{\ \rm per\ cent}}$. The results show that ml-moc is a reliable approach to segregate open-cluster members from field stars.

scholarly journals Hunting for open clusters in Gaia DR2: 582 new open clusters in the Galactic disc

2020 ◽  
Vol 635 ◽  
pp. A45 ◽  
Author(s):  
A. Castro-Ginard ◽  
C. Jordi ◽  
X. Luri ◽  
J. Álvarez Cid-Fuentes ◽  
L. Casamiquela ◽  
...  
Keyword(s):  
Machine Learning ◽  
Big Data ◽  
Clustering Algorithm ◽  
Open Cluster ◽  
Open Clusters ◽  
Galactic Disc ◽  
Data Environment ◽  
Blind Search ◽  
Tidal Tails ◽  
Gaia Dr2

Context. Open clusters are key targets for studies of Galaxy structure and evolution, and stellar physics. Since the Gaia data release 2 (DR2), the discovery of undetected clusters has shown that previous surveys were incomplete. Aims. Our aim is to exploit the Big Data capabilities of machine learning to detect new open clusters in Gaia DR2, and to complete the open cluster sample to enable further studies of the Galactic disc. Methods. We use a machine-learning based methodology to systematically search the Galactic disc for overdensities in the astrometric space and identify the open clusters using photometric information. First, we used an unsupervised clustering algorithm, DBSCAN, to blindly search for these overdensities in Gaia DR2 (l, b, ϖ, μα*, μδ), and then we used a deep learning artificial neural network trained on colour–magnitude diagrams to identify isochrone patterns in these overdensities, and to confirm them as open clusters. Results. We find 582 new open clusters distributed along the Galactic disc in the region |b| < 20°. We detect substructure in complex regions, and identify the tidal tails of a disrupting cluster UBC 274 of ∼3 Gyr located at ∼2 kpc. Conclusions. Adapting the mentioned methodology to a Big Data environment allows us to target the search using the physical properties of open clusters instead of being driven by computational limitations. This blind search for open clusters in the Galactic disc increases the number of known open clusters by 45%.

scholarly journals UOCS*. III. UVIT catalogue of open clusters with machine learning based membership using Gaia EDR3 astrometry

2021 ◽  
Author(s):  
Vikrant V Jadhav ◽  
Clara M Pennock ◽  
Annapurni Subramaniam ◽  
Ram Sagar ◽  
Prasanta Kumar Nayak
Keyword(s):  
Machine Learning ◽  
Main Sequence ◽  
Learning Algorithm ◽  
Ultra Violet ◽  
Gaussian Mixture ◽  
Open Clusters ◽  
Red Giants ◽  
Cluster Membership ◽  
Blue Stragglers ◽  
Red Clump

Abstract We present a study of six open clusters (Berkeley 67, King 2, NGC 2420, NGC 2477, NGC 2682 and NGC 6940) using the Ultra Violet Imaging Telescope (UVIT) aboard ASTROSAT and Gaia EDR3. We used combinations of astrometric, photometric and systematic parameters to train and supervise a machine learning algorithm along with a Gaussian mixture model for the determination of cluster membership. This technique is robust, reproducible and versatile in various cluster environments. In this study, the Gaia EDR3 membership catalogues are provided along with classification of the stars as members, candidates and field in the six clusters. We could detect 200–2500 additional members using our method with respect to previous studies, which helped estimate mean space velocities, distances, number of members and core radii. UVIT photometric catalogues, which include blue stragglers, main-sequence and red giants are also provided. From UV–Optical colour-magnitude diagrams, we found that majority of the sources in NGC 2682 and a few in NGC 2420, NGC 2477 and NGC 6940 showed excess UV flux. NGC 2682 images have ten white dwarf detection in far-UV. The far-UV and near-UV images of the massive cluster NGC 2477 have 92 and 576 members respectively, which will be useful to study the UV properties of stars in the extended turn-off and in various evolutionary stages from main-sequence to red clump. Future studies will carry out panchromatic and spectroscopic analysis of noteworthy members detected in this study.

scholarly journals Membership Lists for 431 Open Clusters in Gaia DR2 Using Extreme Deconvolution Gaussian Mixture Models

2021 ◽  
Vol 923 (1) ◽  
pp. 129
Author(s):  
Karl Jaehnig ◽  
Jonathan Bird ◽  
Kelly Holley-Bockelmann
Keyword(s):  
Mixture Models ◽  
Open Cluster ◽  
Gaussian Mixture Models ◽  
Measurement Data ◽  
Gaussian Mixture ◽  
Open Clusters ◽  
Line Of Sight ◽  
Deconvolution Techniques ◽  
Gaia Dr2 ◽  
Full Covariance Matrix

Abstract Open clusters are groups of stars that form at the same time, making them an ideal laboratory to test theories of star formation, stellar evolution, and dynamics in the Milky Way disk. However, the utility of an open cluster can be limited by the accuracy and completeness of its known members. Here, we employ a “top-down” technique, Extreme Deconvolution Gaussian Mixture Models (XDGMMs), to extract and evaluate known open clusters from Gaia DR2 by fitting the distribution of stellar parallax and proper motion along a line of sight. Extreme deconvolution techniques can recover the intrinsic distribution of astrometric quantities, accounting for the full covariance matrix of the errors; this allows open cluster members to be identified even when presented with relatively uncertain measurement data. To date, open cluster studies have only applied extreme deconvolution to specialized searches for individual systems. We use XDGMMs to characterize the open clusters reported by Ahumada & Lapasset and are able to recover 420 of the 426 open clusters therein (98.1%). Our membership list contains the overwhelming majority (>95%) of previously known cluster members. We also identify a new, significant, and relatively faint cluster member population and validate their membership status using Gaia eDR3. We report the fortuitous discovery of 11 new open cluster candidates within the lines of sight we analyzed. We present our technique, as well as its advantages and challenges, and publish our membership lists and updated cluster parameters.

scholarly journals White dwarf-open cluster associations based on Gaia DR2

2020 ◽  
Vol 645 ◽  
pp. A13
Author(s):  
M. Prišegen ◽  
M. Piecka ◽  
N. Faltová ◽  
M. Kajan ◽  
E. Paunzen
Keyword(s):  
Open Cluster ◽  
Open Clusters ◽  
Photometric Data ◽  
Mass Relation ◽  
Physical Processes ◽  
Satellite Mission ◽  
Cluster Membership ◽  
Fundamental Parameters ◽  
Data Release ◽  
Gaia Dr2

Context. Fundamental parameters and physical processes leading to the formation of white dwarfs (WDs) may be constrained and refined by discovering WDs in open clusters (OCs). Cluster membership can be utilized to establish the precise distances, luminosities, ages, and progenitor masses of such WDs. Aims. We compile a list of probable WDs that are OC members in order to facilitate WD studies that are impractical or difficult to conduct for Galactic field WDs. Methods. We use recent catalogs of WDs and OCs that are based on the second data release of the Gaia satellite mission (GDR2) to identify WDs that are OC members. This crossmatch is facilitated by the astrometric and photometric data contained in GDR2 and the derived catalogs. Assuming that most of the WD members are of the DA type, we estimate the WD masses, cooling ages, and progenitor masses. Results. We have detected several new likely WD members and reassessed the membership of the literature WDs that had been previously associated with the studied OCs. Several of the recovered WDs fall into the recently reported discontinuity in the initial-final mass relation (IFMR) around Mi ∼ 2.0 M⊙, which allows for tighter constrains on the IFMR in this regime.

scholarly journals Orbital Parameters and Binary Properties of 37 FGK Stars in the Cores of Open Clusters NGC 2516 and NGC 2422

2021 ◽  
Vol 162 (6) ◽  
pp. 285
Author(s):  
Isabel Lipartito ◽  
John I. Bailey III ◽  
Timothy D. Brandt ◽  
Benjamin A. Mazin ◽  
Mario Mateo ◽  
...  
Keyword(s):  
Open Cluster ◽  
Open Clusters ◽  
Spectroscopic Binaries ◽  
Proper Motions ◽  
Orbital Eccentricity ◽  
Critical Window ◽  
Cluster Membership ◽  
Orbital Parameters ◽  
Keplerian Orbits ◽  
Cluster Cores

Abstract We present orbits for 24 binaries in the field of open cluster NGC 2516 (∼150 Myr) and 13 binaries in the field of open cluster NGC 2422 (∼130 Myr) using results from a multiyear radial-velocity (RV) survey of the cluster cores. Six of these systems are double-lined spectroscopic binaries. We fit these RV variable systems with orvara, a MCMC-based fitting program that models Keplerian orbits. We use precise stellar parallaxes and proper motions from Gaia EDR3 to determine cluster membership. We impose a barycentric RV prior on all cluster members; this significantly improves our orbital constraints. Two of our systems have periods between five and 15 days, the critical window in which tides efficiently damp orbital eccentricity. These binaries should be included in future analyses of circularization across similarly-aged clusters. We also find a relatively flat distribution of binary mass ratios, consistent with previous work. With the inclusion of TESS light curves for all available targets, we identity target 378–036252 as a new eclipsing binary. We also identify a field star whose secondary has a mass in the brown dwarf range, as well as two cluster members whose RVs suggest the presence of an additional companion. Our orbital fits will help constrain the binary fraction and binary properties across stellar age and across stellar environment.

scholarly journals Reanalysis of nearby open clusters using Gaia DR1/TGAS and HSOY

2018 ◽  
Vol 615 ◽  
pp. A12 ◽  
Author(s):  
Steffi X. Yen ◽  
Sabine Reffert ◽  
Elena Schilbach ◽  
Siegfried Röser ◽  
Nina V. Kharchenko ◽  
...  
Keyword(s):  
Milky Way ◽  
Open Cluster ◽  
Star Clusters ◽  
Open Clusters ◽  
Parameter Estimates ◽  
Cluster Membership ◽  
Fitting Procedure ◽  
Cluster Data ◽  
The Galaxy ◽  
Automatic Tool

Context. Open clusters have long been used to gain insights into the structure, composition, and evolution of the Galaxy. With the large amount of stellar data available for many clusters in the Gaia era, new techniques must be developed for analyzing open clusters, as visual inspection of cluster color-magnitude diagrams is no longer feasible. An automatic tool will be required to analyze large samples of open clusters. Aims. We seek to develop an automatic isochrone-fitting procedure to consistently determine cluster membership and the fundamental cluster parameters. Methods. Our cluster characterization pipeline first determined cluster membership with precise astrometry, primarily from TGAS and HSOY. With initial cluster members established, isochrones were fitted, using a χ2 minimization, to the cluster photometry in order to determine cluster mean distances, ages, and reddening. Cluster membership was also refined based on the stellar photometry. We used multiband photometry, which includes ASCC-2.5 BV, 2MASS JHKs, and Gaia G band. Results. We present parameter estimates for all 24 clusters closer than 333 pc as determined by the Catalogue of Open Cluster Data and the Milky Way Star Clusters catalog. We find that our parameters are consistent to those in the Milky Way Star Clusters catalog. Conclusions. We demonstrate that it is feasible to develop an automated pipeline that determines cluster parameters and membership reliably. After additional modifications, our pipeline will be able to use Gaia DR2 as input, leading to better cluster memberships and more accurate cluster parameters for a much larger number of clusters.

scholarly journals A Gaia DR2 view of the open cluster population in the Milky Way

2018 ◽  
Vol 618 ◽  
pp. A93 ◽  
Author(s):  
T. Cantat-Gaudin ◽  
C. Jordi ◽  
A. Vallenari ◽  
A. Bragaglia ◽  
L. Balaguer-Núñez ◽  
...  
Keyword(s):  
Stellar Evolution ◽  
Milky Way ◽  
Galactic Plane ◽  
Galactic Disk ◽  
Open Cluster ◽  
Open Clusters ◽  
Number Of Clusters ◽  
History Of ◽  
Gaia Dr2

Context. Open clusters are convenient probes of the structure and history of the Galactic disk. They are also fundamental to stellar evolution studies. The second Gaia data release contains precise astrometry at the submilliarcsecond level and homogeneous photometry at the mmag level, that can be used to characterise a large number of clusters over the entire sky. Aims. In this study we aim to establish a list of members and derive mean parameters, in particular distances, for as many clusters as possible, making use of Gaia data alone. Methods. We compiled a list of thousands of known or putative clusters from the literature. We then applied an unsupervised membership assignment code, UPMASK, to the Gaia DR2 data contained within the fields of those clusters. Results. We obtained a list of members and cluster parameters for 1229 clusters. As expected, the youngest clusters are seen to be tightly distributed near the Galactic plane and to trace the spiral arms of the Milky Way, while older objects are more uniformly distributed, deviate further from the plane, and tend to be located at larger Galactocentric distances. Thanks to the quality of Gaia DR2 astrometry, the fully homogeneous parameters derived in this study are the most precise to date. Furthermore, we report on the serendipitous discovery of 60 new open clusters in the fields analysed during this study.

scholarly journals Characterizing dynamical stages of open clusters located in the Sagittarius spiral arm

2019 ◽  
Vol 488 (2) ◽  
pp. 1635-1651 ◽  
Author(s):  
M S Angelo ◽  
A E Piatti ◽  
W S Dias ◽  
F F S Maia
Keyword(s):  
Structural Parameters ◽  
Three Dimensional ◽  
Dynamical Evolution ◽  
Open Clusters ◽  
Mass Star ◽  
Proper Motions ◽  
Concentration Parameter ◽  
Low Mass ◽  
Gaia Dr2 ◽  
Spiral Arm

Abstract The study of dynamical properties of Galactic open clusters (OCs) is a fundamental prerequisite for the comprehension of their dissolution processes. In this work, we characterized 12 OCs, namely: Collinder 258, NGC 6756, Czernik 37, NGC 5381, Ruprecht 111, Ruprecht 102, NGC 6249, Basel 5, Ruprecht 97, Trumpler 25, ESO 129−SC32, and BH 150, projected against dense stellar fields. In order to do that, we employed Washington CT1 photometry and Gaia DR2 astrometry, combined with a decontamination algorithm applied to the three-dimensional astrometric space of proper motions and parallaxes. From the derived membership likelihoods, we built decontaminated colour–magnitude diagrams, while structural parameters were obtained from King profiles fitting. Our analysis revealed that they are relatively young OCs (log(t  yr−1) ∼7.3–8.6), placed along the Sagittarius spiral arm, and at different internal dynamical stages. We found that the half-light radius to Jacobi radius ratio, the concentration parameter and the age to relaxation time ratio describe satisfactorily their different stages of dynamical evolution. Those relative more dynamically evolved OCs have apparently experienced more important low-mass star loss.

scholarly journals A Cooperative Machine Learning Approach for Pedestrian Navigation in Indoor IoT

Sensors ◽  
2019 ◽  
Vol 19 (21) ◽  
pp. 4609 ◽  
Author(s):  
Marzieh Jalal Abadi ◽  
Luca Luceri ◽  
Mahbub Hassan ◽  
Chun Tung Chou ◽  
Monica Nicoli
Keyword(s):  
Machine Learning ◽  
A Priori ◽  
Dead Reckoning ◽  
Step Length ◽  
Consensus Algorithm ◽  
Learning Approach ◽  
Indoor Environments ◽  
Machine Learning Approach ◽  
Multiple Devices ◽  
Priori Information

This paper presents a system based on pedestrian dead reckoning (PDR) for localization of networked mobile users, which relies only on sensors embedded in the devices and device- to-device connectivity. The user trajectory is reconstructed by measuring step by step the user displacements. Though step length can be estimated rather accurately, heading evaluation is extremely problematic in indoor environments. Magnetometer is typically used, however measurements are strongly perturbed. To improve the location accuracy, this paper proposes a novel cooperative system to estimate the direction of motion based on a machine learning approach for perturbation detection and filtering, combined with a consensus algorithm for performance augmentation by cooperative data fusion at multiple devices. A first algorithm filters out perturbed magnetometer measurements based on a-priori information on the Earth’s magnetic field. A second algorithm aggregates groups of users walking in the same direction, while a third one combines the measurements of the aggregated users in a distributed way to extract a more accurate heading estimate. To the best of our knowledge, this is the first approach that combines machine learning with consensus algorithms for cooperative PDR. Compared to other methods in the literature, the method has the advantage of being infrastructure-free, fully distributed and robust to sensor failures thanks to the pre-filtering of perturbed measurements. Extensive indoor experiments show that the heading error is highly reduced by the proposed approach thus leading to noticeable enhancements in localization performance.

scholarly journals The nature of the NGC 2546: Not one but two open clusters

2020 ◽  
Vol 633 ◽  
pp. A146 ◽  
Author(s):  
A. D. Alejo ◽  
J. F. González ◽  
M. E. Veramendi
Keyword(s):  
Radial Velocity ◽  
Binary Stars ◽  
Open Cluster ◽  
Open Clusters ◽  
Spectroscopic Binaries ◽  
Physical Parameters ◽  
Distance Modulus ◽  
Cluster Membership ◽  
Spectroscopic Observations ◽  
Astrometric Data

Context. As part of a broader project on the role of binary stars in clusters, we present a spectroscopic study of the open cluster NGC 2546, which is a large cluster lacking previous spectroscopic analysis. Aims. We report the finding of two open clusters in the region of NGC 2546. For the two star groups, we determine radial velocity, parallax, proper motion, reddening, distance modulus, and age, using our spectroscopic observations and available photometric and astrometric data, mainly from the second Gaia data release (Gaia-DR2). We also determine the orbit of four spectroscopic binaries in these open clusters. Methods. From mid-resolution spectroscopic observations for 28 stars in the NGC 2546 region, we determined radial velocities and evaluate velocity variability. To analyze double-lined spectroscopic binaries, we used a spectral separation technique and fit the spectroscopic orbits using a least-squares code. The presence of two stellar groups is suggested by the radial velocity distribution and confirmed by available photometric and astrometric data. We applied a multi-criteria analysis to determine cluster membership, and obtained kinematic and physical parameters of the clusters. Results. NGC 2546 is actually two clusters, NGC 2546A and NGC 2546B, which are not physically related to each other. NGC 2546A has an age of about 180 Myr and a distance of 950 pc. It has a half-number radius of 8 pc and contains about 480 members brighter than G = 18 mag. NGC 2546B is a very young cluster (<10 Myr) located at a distance of 1450 pc. It is a small cluster with 80 members and a half-number radius of 1.6 pc. Stars less massive than 2.5 M⊙ in this cluster would be pre-main-sequence objects. We detected four spectroscopic binaries and determined their orbits. The two binaries of NGC 2546A contain chemically peculiar components: HD 68693 is composed of two mercury-manganese stars and HD 68624 has a Bp silicon secondary. Among the most massive objects of NGC 2546B, there are two binary stars: HD 68572, with P = 124.2 d, and CD -37 4344 with P = 10.4 d.

Sign in / Sign up

Export Citation Format

Share Document