Search

Start Over Subject classification Content Level Research
11. YSO candidate catalog from ANN
ID:
ivo://CDS.VizieR/J/A+A/647/A116
Title:
YSO candidate catalog from ANN
Short Name:
J/A+A/647/A116
Date:
21 Oct 2021
Publisher:
CDS
Description:
Observed young stellar objects (YSOs) are used to study star formation and characterize star-forming regions. For this purpose, YSO candidate catalogs are compiled from various surveys, especially in the infrared (IR), and simple selection schemes in color-magnitude diagrams (CMDs) are often used to identify and classify YSOs. We propose a methodology for YSO classification through machine learning (ML) using Spitzer IR data. We detail our approach in order to ensure reproducibility and provide an in-depth example on how to efficiently apply ML to an astrophysical classification. We used feed forward artificial neural networks (ANNs) that use the four IRAC bands (3.6, 4.5, 5.8, and 8 micron) and the 24 micron MIPS band from Spitzer to classify point source objects into CI and CII YSO candidates or as contaminants. We focused on nearby (~1kpc) star-forming regions including Orion and NGC 2264, and assessed the generalization capacity of our network from one region to another. We found that ANNs can be efficiently applied to YSO classification with a contained number of neurons (~25). Knowledge gathered on one star-forming region has shown to be partly efficient for prediction in new regions. The best generalization capacity was achieved using a combination of several star-forming regions to train the network. Carefully rebalancing the training proportions was necessary to achieve good results. We observed that the predicted YSOs are mainly contaminated by under-constrained rare subclasses like Shocks and polycyclic aromatic hydrocarbons (PAHs), or by the vastly dominant other kinds of stars (mostly on the main sequence). We achieved above 90% and 97% recovery rate for CI and CII YSOs, respectively, with a precision above 80% and 90% for our most general results. We took advantage of the great flexibility of ANNs to define, for each object, an effective membership probability to each output class. Using a threshold in this probability was found to efficiently improve the classification results at a reasonable cost of object exclusion. With this additional selection, we reached 90% and 97% precision on CI and CII YSOs, respectively, for more than half of them. Our catalog of YSO candidates in Orion (365 CI, 2381 CII) and NGC 2264 (101 CI, 469 CII) predicted by our final ANN, along with the class membership probability for each object, is publicly available at the CDS. Compared to usual CMD selection schemes, ANNs provide a possibility to quantitatively study the properties and quality of the classification. Although some further improvement may be achieved by using more powerful ML methods, we established that the result quality depends mostly on the training set construction. Improvements in YSO identification with IR surveys using ML would require larger and more reliable training catalogs, either by taking advantage of current and future surveys from various facilities like VLA, ALMA, or Chandra, or by synthesizing such catalogs from simulations.
12. YSOs in California Molecular Cloud
ID:
ivo://CDS.VizieR/J/A+A/606/A100
Title:
YSOs in California Molecular Cloud
Short Name:
J/A+A/606/A100
Date:
21 Oct 2021
Publisher:
CDS
Description:
We present new high resolution and dynamic range dust column density and temperature maps of the California Molecular Cloud derived from a combination of Planck and Herschel dust-emission maps, and 2MASS NIR dust-extinction maps. We used these data to determine the ratio of the 2.2{mu}m extinction coefficient to the 850{mu}m opacity and found the value to be close to that found in similar studies of the Orion B and Perseus clouds but higher than that characterizing the Orion A cloud, indicating that variations in the fundamental optical properties of dust may exist between local clouds. We show that over a wide range of extinction, the column density probability distribution function (pdf) of the cloud can be well described by a simple power law (i.e., PDF_N_{prop.to}A_K_^-n^) with an index (n=4.0+/-0.1) that represents a steeper decline with A_K_ than found (n~=3) in similar studies of the Orion and Perseus clouds. Using only the protostellar population of the cloud and our extinction maps we investigate the Schmidt relation, that is, the relation between the protostellar surface density, {Sigma}_*_, and extinction, A_K_, within the cloud. We show that {Sigma}_*_ is directly proportional to the ratio of the protostellar and cloud pdfs, i.e., PDF_*_(A_K_)/PDF_N_(A_K_). We use the cumulative distribution of protostars to infer the functional forms for both {Sigma}_*_ and PDF_*_. We find that {Sigma}_*_ is best described by two power-law functions. At extinctions A_K_<=2.5mag, {Sigma}_*_{prop.to}A_K_^{beta}^ with {beta}=3.3 while at higher extinctions {beta}=2.5, both values steeper than those (~=2) found in other local giant molecular clouds (GMCs). We find that PDF_*_ is a declining function of extinction also best described by two power-laws whose behavior mirrors that of {Sigma}_*_. Our observations suggest that variations both in the slope of the Schmidt relation and in the sizes of the protostellar populations between GMCs are largely driven by variations in the slope, n, of PDF_N_(A_K_). This confirms earlier studies suggesting that cloud structure plays a major role in setting the global star formation rates in GMCs