اشترك بالحزمة الذهبية واحصل على وصول غير محدود شمرا أكاديميا
تسجيل مستخدم جديدMorphological classification of galaxies with deep learning: comparing 3-way and 4-way CNNs
73
0
0.0
(
0
)
اسأل ChatGPT حول البحث
ﻻ يوجد ملخص باللغة العربية
Classifying the morphologies of galaxies is an important step in understanding their physical properties and evolutionary histories. The advent of large-scale surveys has hastened the need to develop techniques for automated morphological classification. We train and test several convolutional neural network architectures to classify the morphologies of galaxies in both a 3-class (elliptical, lenticular, spiral) and 4-class (+irregular/miscellaneous) schema with a dataset of 14034 visually-classified SDSS images. We develop a new CNN architecture that outperforms existing models in both 3 and 4-way classification, with overall classification accuracies of 83% and 81% respectively. We also compare the accuracies of 2-way / binary classifications between all four classes, showing that ellipticals and spirals are most easily distinguished (>98% accuracy), while spirals and irregulars are hardest to differentiate (78% accuracy). Through an analysis of all classified samples, we find tentative evidence that misclassifications are physically meaningful, with lenticulars misclassified as ellipticals tending to be more massive, among other trends. We further combine our binary CNN classifiers to perform a hierarchical classification of samples, obtaining comparable accuracies (81%) to the direct 3-class CNN, but considerably worse accuracies in the 4-way case (65%). As an additional verification, we apply our networks to a small sample of Galaxy Zoo images, obtaining accuracies of 92%, 82% and 77% for the binary, 3-way and 4-way classifications respectively.
قيم البحث
اقرأ أيضاً
We present an extended morphometric system to automatically classify galaxies from astronomical images. The new system includes the original and modifie
There are several supervised machine learning methods used for the application of automated morphological classification of galaxies; however, there has not yet been a clear comparison of these different methods using imaging data, or a investigation
for maximising their effectiveness. We carry out a comparison between several common machine learning methods for galaxy classification (Convolutional Neural Network (CNN), K-nearest neighbour, Logistic Regression, Support Vector Machine, Random Forest, and Neural Networks) by using Dark Energy Survey (DES) data combined with visual classifications from the Galaxy Zoo 1 project (GZ1). Our goal is to determine the optimal machine learning methods when using imaging data for galaxy classification. We show that CNN is the most successful method of these ten methods in our study. Using a sample of $sim$2,800 galaxies with visual classification from GZ1, we reach an accuracy of $sim$0.99 for the morphological classification of Ellipticals and Spirals. The further investigation of the galaxies that have a different ML and visual classification but with high predicted probabilities in our CNN usually reveals an the incorrect classification provided by GZ1. We further find the galaxies having a low probability of being either spirals or ellipticals are visually Lenticulars (S0), demonstrating that supervised learning is able to rediscover that this class of galaxy is distinct from both Es and Spirals. We confirm that $sim$2.5% galaxies are misclassified by GZ1 in our study. After correcting these galaxies labels, we improve our CNN performance to an average accuracy of over 0.99 (accuracy of 0.994 is our best result).
Galaxy morphology is a fundamental quantity, that is essential not only for the full spectrum of galaxy-evolution studies, but also for a plethora of science in observational cosmology. While a rich literature exists on morphological-classification t
echniques, the unprecedented data volumes, coupled, in some cases, with the short cadences of forthcoming Big-Data surveys (e.g. from the LSST), present novel challenges for this field. Large data volumes make such datasets intractable for visual inspection (even via massively-distributed platforms like Galaxy Zoo), while short cadences make it difficult to employ techniques like supervised machine-learning, since it may be impractical to repeatedly produce training sets on short timescales. Unsupervised machine learning, which does not require training sets, is ideally suited to the morphological analysis of new and forthcoming surveys. Here, we employ an algorithm that performs clustering of graph representations, in order to group image patches with similar visual properties and objects constructed from those patches, like galaxies. We implement the algorithm on the Hyper-Suprime-Cam Subaru-Strategic-Program Ultra-Deep survey, to autonomously reduce the galaxy population to a small number (160) of morphological clusters, populated by galaxies with similar morphologies, which are then benchmarked using visual inspection. The morphological classifications (which we release publicly) exhibit a high level of purity, and reproduce known trends in key galaxy properties as a function of morphological type at z<1 (e.g. stellar-mass functions, rest-frame colours and the position of galaxies on the star-formation main sequence). Our study demonstrates the power of unsupervised machine learning in performing accurate morphological analysis, which will become indispensable in this new era of deep-wide surveys.
We present the star formation histories (SFHs) of 20 faint M31 satellites ($-12 lesssim M_V lesssim -6$) that were measured by modeling sub-horizontal branch (HB) depth color-magnitude diagrams constructed from Hubble Space Telescope (HST) imaging. R
einforcing previous results, we find that virtually all galaxies quenched between 3 and 9 Gyr ago, independent of luminosity, with a notable concentration $3-6$ Gyr ago. This is in contrast to the Milky Way (MW) satellites, which are generally either faint with ancient quenching times or luminous with recent ($<3$ Gyr) quenching times. We suggest that systematic differences in the quenching times of M31 and MW satellites may be a reflection of the varying accretion histories of M31 and the MW. This result implies that the formation histories of low-mass satellites may not be broadly representative of low-mass galaxies in general. Among the M31 satellite population we identify two distinct groups based on their SFHs: one with exponentially declining SFHs ($tau sim 2$ Gyr) and one with rising SFHs with abrupt quenching. We speculate how these two groups could be related to scenarios for a recent major merger involving M31. The Cycle 27 HST Treasury survey of M31 satellites will provide well-constrained ancient SFHs to go along with the quenching times we measure here. The discovery and characterization of M31 satellites with $M_V gtrsim -6$ would help quantify the relative contributions of reionization and environment to quenching of the lowest-mass satellites.
We study the two main constituent galaxies of a constrained simulation of the Local Group as candidates for the Milky Way (MW) and Andromeda (M31). We focus on the formation of the stellar discs and its relation to the formation of the group as a ric
h system with two massive galaxies, and investigate the effects of mergers and accretion as drivers of morphological transformations. We use a state-of-the-art hydrodynamical code which includes star formation, feedback and chemical enrichment to carry out our study. We run two simulations, where we include or neglect the effects of radiation pressure from stars, to investigate the impact of this process on the morphologies and star formation rates of the simulated galaxies. We find that the simulated M31 and MW have different formation histories, even though both inhabit, at z=0, the same environment. These differences directly translate into and explain variations in their star formation rates, in-situ fractions and final morphologies. The M31 candidate has an active merger history, as a result of which its stellar disc is unable to survive unaffected until the present time. In contrast, the MW candidate has a smoother history with no major mergers at late times, and forms a disc that grows steadily; at z=0 the simulated MW has an extended, rotationally-supported disc which is dominant over the bulge. Our two feedback implementations predict similar evolution of the galaxies and their discs, although some variations are detected, the most important of which is the formation time of the discs: in the model with weaker/stronger feedback the discs form earlier/later. In summary, by comparing the formation histories of the two galaxies, we conclude that the particular merger/accretion history of a galaxy rather than its environment at the LG-scales is the main driver of the formation and subsequent growth or destruction of galaxy discs.
سجل دخول لتتمكن من نشر تعليقات
التعليقات
جاري جلب التعليقات
سجل دخول لتتمكن من متابعة معايير البحث التي قمت باختيارها
