基于EfficientNet的星系形态分类研究

星系的结构和形态能够反映星系自身的物理性质,其形态的分类是后续分析研究的一个重要环节.EfficientNet模型使用复合系数对深度网络模型的深度、宽度、输入图像分辨率进行更加结构化的统一缩放,是一种新的深度网络优化扩展方法.将该模型应用于星系数据形态的分类研究中,结果表明基于EfficientNetB5模型的平均准确率、精确率、召回率以及F1分数(精确率与召回率的调和平均数)都在96.6%以上,与残差网络(Residual network, ResNet)中ResNet-26模型的分类结果相比有较大的提升.实验结果证明EfficientNet的深度网络优化扩展方法可行且有效,可应用于星系的形态分类.     

基于SE-Inception-v3的星系形态分类模型

随着天文探测技术的快速发展, 海量的星系图像数据不断产生, 能够及时高效地对星系图像进行形态分类对研究星系的形成与演化至关重要. 针对传统的星系形态分类模型特征选择困难、分类速度慢、准确率受限等难题, 提出一种以Inception-v3神经网络为主干结构, 融合压缩激励(Squeeze and Excitation Network, SE)通道注意力机制的星系形态分类模型. 该模型在斯隆数字巡天(Sloan Digital Sky Survey, SDSS)样本的测试集准确率高达99.37%. 旋涡星系、圆形星系、中间星系、雪茄状星系与侧向星系的F1值分别为99.33%、99.58%、99.33%、99.41%与99.16%. 该模型与Inception-v3、MobileNet (Mobile Neural Network)和ResNet (Residual Neural Network)网络模型相比, SE-Inception-v3宽度和深度优势表现出更强的特征提取能力, 可以高效识别不同形态的星系, 为未来大型巡天计划的大规模星系形态分类问题提供了一种新方法.     

EfficientNetV2-S-Triplet7: 一种改进的星系newline 形态学分类算法

星系的形态与星系的形成和演化息息相关, 其形态学分类是星系天文学后续研究的重要一环. 当前海量天文观测数据的出现使得天文数据自动分析方法越来越得到重视, 针对此问题, 利用先进的深度学习骨干网络EfficientNetV2, 分析不同的注意力机制类型和使用节点对网络性能的影响, 构建了一种命名为EfficientNetV2-S-Triplet7 (即在EfficientNetV2-S stage7的$1\times1$卷积层后加入Triplet模块)的改进算法模型来实现星系形态学的自动分类. 使用第二期星系动物园(Galaxy Zoo 2, GZ2)中超过24万张的测光图像作为初始数据进行实验测试. 在对数据进行预处理时采取了尺寸抖动、翻转、色彩畸变等图像增强手段来解决图像数量的不平衡问题. 在同一系列经典和前沿的深度学习算法模型AlexNet、ResNet-34、MobileNetV2、RegNet进行对比实验后, 得出EfficientNetV2-S-Triplet7算法在分类准确率、查全率和F1分数等指标上具有最好的测试结果. 在9375张测试图像中的3项指标值分别可达到89.03%、90.21%、89.93%, 查准率达到89.69%, 在其他模型中排在第3位. 该结果表明将EfficientNetV2-S-Triplet7算法应用于大规模星系数据的形态学分类任务中有很好的效果.     

基于深度残差网络的星系形态分类

星系形态与星系的形成和演化有着密切的联系,因此星系形态分类(galaxy morphology classification)成为研究不同星系物理特征的重要过程之一。斯隆数字巡天(Sloan Digital Sky Survey, SDSS)等大型巡天计划产生的海量星系图像数据对星系形态的准确、实时分类提出了新的挑战,而深度学习(deep learning)算法能有效应对这类海量星系图片的自动分类考验。面向星系形态分类问题提出了一种改进的深度残差网络(residual network, ResNet),即ResNet-26模型。该模型对残差单元进行改进,减少了网络深度,并增加了网络宽度,实现了对星系形态特征的自动提取、识别和分类。实验结果表明,与Dieleman和ResNet-50等其他流行的卷积神经网络(convolution neural network, CNN)模型相比,ResNet-26模型具有更优的分类性能,可应用于未来大型巡天计划的大规模星系形态分类系统。     

基于Stacking集成学习的恒星/星系分类研究

机器学习在当今诸多领域已经取得了巨大的成功,但是机器学习的预测效果往往依赖于具体问题.集成学习通过综合多个基分类器来预测结果,因此,其适应各种场景的能力较强,分类准确率较高.基于斯隆数字巡天(Sloan Digital Sky Survey,SDSS)计划恒星/星系中最暗源星等集分类正确率低的问题,提出一种基于Stacking集成学习的恒星/星系分类算法.从SDSS-DR7(SDSS Data Release 7)中获取完整的测光数据集,并根据星等值划分为亮源星等集、暗源星等集和最暗源星等集.仅针对分类较为复杂且困难的最暗源星等集展开分类研究.首先,对最暗源星等集使用10折嵌套交叉验证,然后使用支持向量机(Support Vector Machine,SVM)、随机森林(Random Forest,RF)、XGBoost(eXtreme Gradient Boosting)等算法建立基分类器模型;使用梯度提升树(Gradient Boosting Decision Tree,GBDT)作为元分类器模型.最后,使用基于星系的分类正确率等指标,与功能树(Function Tree,FT)、SVM、RF、GBDT、XGBoost、堆叠降噪自编码(Stacked Denoising AutoEncoders,SDAE)、深度置信网络(Deep Belief Network,DBN)、深度感知决策树(Deep Perception Decision Tree,DPDT)等模型进行分类结果对比分析.实验结果表明,Stacking集成学习模型在最暗源星等集分类中要比FT算法的星系分类正确率提高了将近10%.同其他传统的机器学习算法、较强的提升算法、深度学习算法相比,Stacking集成学习模型也有较大的提升.     

早型旋涡星系M81 (NGC 3031)的形态学解构研究

利用星系解构软件GALFIT通过面亮度轮廓拟合对近邻早型旋涡星系M81 (NGC 3031)进行形态学解构,旨在探究M81星系的结构组成并对其进行形态学量化.通过6种解构模式,对M81进行了不同复杂程度的结构分解,其中最复杂的解构模式包含核球、盘、外旋臂、内旋臂、星系核5个子结构.研究结果显示, M81有一个Sérsic指数约为5.0的经典核球,其形态和光度在不同解构模式中均保持稳定; M81星系盘的Sérsic指数约为1.2,但它的形态参数和光度与是否分解内旋臂相关.不同子结构的组合对作为混合体的星系整体的形态有不可忽视的影响.星系解构的结果提供了不同解构模式适用性的建议:其中核球+盘+星系核的三成分解构适用于大样本星系的核-盘研究;而考虑旋臂的复杂解构则适合于对星系子结构的精确测量,如小样本(或个源)研究.基于Spitzer-The Infrared Array Camera (IRAC) 4.5μm的单波段图像的形态学解构研究是后续一系列研究的开始,在此基础上未来将会对M81进行多波段解构,同时研究不同子结构的光谱能量分布和星族性质,并推断M81各子结构的形成历史和演化过程.     

COSMOS场中高红移大质量星系的形态分类研究

基于COSMOS(Cosmic Evolution Survey)天区的多波段测光数据和HST(Hubble Space Telescope)近红外高分辨率观测图像,利用质量限(恒星质量M*≥1010.5M⊙)选取了362个红移分布在1≤z≤3的星系样本,并对这些大质量星系的形态特征进行了分类研究.来自UVJ(U-V和V-J)双色图分类系统、目视分类系统、非模型化分类系统(基尼系数G和矩指数M20)和模型化分类系统(S′ersic index,n)的分类结果彼此相一致.相比较于恒星形成星系(SFGs),通过UVJ双色图定义的宁静星系(QGs)表现出致密的椭圆结构,而且G和n值偏大,但M20和星系有效半径(re)偏小.不同星系分类系统(双色图分类系统、非模型化分类系统和模型化分类系统)定义的SFGs和QGs样本,都明显存在星系的大小随红移的演化关系,这种演化趋势QGs比SFGs更剧烈,而且不依赖于星系分类方法的选择.     

星系团Abell 2199的恒星形成性质

通过对近邻星系团Abell 2199中290颗成员星系进行形态分类,研究这些星系的恒星形成率及其与形态和相关物理特性之间的关系.该星系团中星系的特征恒星形成率与Ha等值宽度、星系光谱在4000A处的跃变程度以及星系所包含的恒星质量之间有较强的相关性.这些星系的恒星形成活动没有表现出明显的环境效应,表明该星系团仍处在剧烈的动力学演化阶段,远没有达到动力学平衡.     

COSMOS场中星系恒星形成的演化研究

基于COSMOS(Cosmic Evolution Survey)/Ultra VISTA(Ultra-deep Visible and Infrared Survey Telescope for Astronomy)场中多波段测光数据,利用质量限选取了红移分布在0z3.5的星系样本.通过UVJ(U-V和V-J)双色图分类判据将星系分类成恒星形成星系(SFGs)和宁静星系(QGs).对于红移分布在0z1.5范围内且M*1011M⊙的QGs来说,该星系在样本中所占比例高于70%.在红移0z3.5范围内,恒星形成星系的恒星形成率(SFR)与恒星质量(M*)之间有着很强的主序(MS)关系.对于某一固定的恒星质量M*来说,星系的SFR和比恒星形成率(s SFR)会随着红移增大而增大,这表明在高红移处恒星形成星系更加活跃,有激烈的恒星形成.相对于低质量的星系来说,高质量的SFGs有较低的s SFR,这意味着低质量星系的增长更多的是通过星系本身的恒星形成.通过结合来自文献中数据点信息,发现更高红移(2z8)星系的s SFR随红移的演化趋势变弱,其演化关系是s SFR∝(1+z)0.94±0.17.     

一种新的基于2维傅里叶谱图像的恒星光谱特征提取方法和深度网络分类应用

天体光谱分类是天文学研究的重要内容之一,其关键是从光谱数据中选择和提取对分类识别最有效的特征构建特征空间.提出一种新的基于2维傅里叶谱图像的特征提取方法,并应用于LAMOST (the Large Sky Area Multi-Object Fiber Spectroscopic Telescope)恒星光谱数据的分类研究中.光谱数据来源于LAMOST Data Release 5(DR5),选取30000条F、 G和K型星光谱数据,利用短时傅里叶变换(Short-Time Fourier Transform, STFT)将1维光谱数据变换成2维傅里叶谱图像,对得到的2维傅里叶谱图像采用深度卷积网络模型进行分类,得到的分类准确率是92.90%.实验结果表明通过对LAMOST恒星光谱数据进行STFT可得到光谱的2维傅里叶谱图像,谱图像构成了新的光谱数据特征和特征空间,新的特征对于光谱数据分类是有效的.此方法是对光谱分类的一种全新尝试,对海量天体光谱的分类和挖掘处理有一定的开创意义.     

Study of Star/Galaxy Classification Based on the XGBoost Algorithm

Machine learning has achieved great success in many areas today. The lifting algorithm has a strong ability to adapt to various scenarios with a high accuracy, and has played a great role in many fields. But in astronomy, the application of lifting algorithms is still rare. In response to the low classification accuracy of the dark star/galaxy source set in the Sloan Digital Sky Survey (SDSS), a new research result of machine learning, eXtreme Gradient Boosting (XGBoost), has been introduced. The complete photometric data set is obtained from the SDSS-DR7, and divided into a bright source set and a dark source set according to the star magnitude. Firstly, the ten-fold cross-validation method is used for the bright source set and the dark source set respectively, and the XGBoost algorithm is used to establish the star/galaxy classification model. Then, the grid search and other methods are used to adjust the XGBoost parameters. Finally, based on the galaxy classification accuracy and other indicators, the classification results are analyzed, by comparing with the models of function tree (FT), Adaptive boosting (Adaboost), Random Forest (RF), Gradient Boosting Decision Tree (GBDT), Stacked Denoising AutoEncoders (SDAE), and Deep Belief Nets (DBN). The experimental results show that, the XGBoost improves the classification accuracy of galaxies in the dark source classification by nearly 10% as compared to the function tree algorithm, and improves the classification accuracy of sources with the darkest magnitudes in the dark source set by nearly 5% as compared to the function tree algorithm. Compared with other traditional machine learning algorithms and deep neural networks, the XGBoost also has different degrees of improvement.     

Research on Star/Galaxy Classification Based on Stacking Ensemble Learning

Machine learning has achieved great success in many areas today, but the forecast effect of machine learning often depends on the specific problem. An ensemble learning forecasts results by combining multiple base classifiers. Therefore, its ability to adapt to various scenarios is strong, and the classification accuracy is high. In response to the low classification accuracy of the darkest source magnitude set of stars/galaxies in the Sloan Digital Sky Survey (SDSS), a star/galaxy classification algorithm based on the stacking ensemble learning is proposed in this paper. The complete photometric data set is obtained from the SDSS Data Release (DR) 7, and divided into the bright source magnitude set, dark source magnitude set, and darkest source magnitude set according to the stellar magnitude. Firstly, the 10-fold nested cross-validation method is used for the darkest source magnitude set, then the Support Vector Machine (SVM), Random Forest (RF), and eXtreme Gradient Boosting (XGBoost) algorithms are used to establish the base-classifier model; the Gradient Boosting Decision Tree (GBDT) is used as the meta-classifier model. Finally, based on the classification accuracy of galaxies and other indicators, the classification results are analyzed and compared with the results obtained by the Function Tree (FT), SVM, RF, GBDT, Stacked Denoising Autoencoders (SDAE), Deep Belief Nets (DBN), and Deep Perception Decision Tree (DPDT) models. The experimental results show that the stacking ensemble learning model has improved the classification accuracy of galaxies in the darkest source magnitude set by nearly 10% compared to the function tree algorithm. Compared with other traditional machine learning algorithm, stronger lifting algorithm, and deep learning algorithm, the stacking ensemble learning model also has different degrees of improvement.     

Dust-penetrated arm classes: insights from rising and falling rotation curves

In the last decade, near-infrared imaging has highlighted the decoupling of gaseous and old stellar discs: the morphologies of optical (Population I) tracers compared to the old stellar disc morphology, can be radically different. Galaxies which appear multi-armed and even flocculent in the optical may show significant grand-design spirals in the near-infrared. Furthermore, the optically determined Hubble classification scheme does not provide a sound way of classifying dust-penetrated stellar discs: spiral arm pitch angles (when measured in the near-infrared) do not correlate with Hubble type. The dust-penetrated classification scheme of Block & Puerari provides an alternative classification based on near-infrared morphology, which is thus more closely linked to the dominant stellar mass component. Here we present near-infrared K -band images of 14 galaxies, on which we have performed a Fourier analysis of the spiral structure in order to determine their near-infrared pitch angles and dust-penetrated arm classes. We have also used the rotation curve data of Mathewson et al. to calculate the rates of shear in the stellar discs of these galaxies. We find a correlation between near-infrared pitch angle and rate of shear: galaxies with wide open arms (the γ class) are found to have rising rotation curves, while those with falling rotation curves belong to the tightly wound α bin. The major determinant of near-infrared spiral arm pitch angle is the distribution of matter within the galaxy concerned. The correlation reported in this study provides the physical basis underpinning spiral arm classes in the dust-penetrated regime and underscores earlier spectroscopic findings by Burstein and Rubin that Hubble type and mass distributions are unrelated.     

Structure through colour: a pixel approach towards understanding galaxies

We present a study of pixel colour–magnitude diagrams (pCMDs) for a sample of 69 nearby galaxies chosen to span a wide range of Hubble types. Our goal is to determine how useful a pixel approach is for studying galaxies according to their stellar light distributions and content. The galaxy images were analysed on a pixel-by-pixel basis to reveal the structure of the individual pCMDs. We find that the average surface brightness (or projected mass density) in each pixel varies according to galaxy type. Early-type galaxies exhibit a clear 'prime sequence' and some pCMDs of face-on spirals reveal 'inverse-L' structures. We find that the colour dispersion at a given magnitude is found to be approximately constant in early-type galaxies but this quantity varies in the mid and late types. We investigate individual galaxies and find that the pCMDs can be used to pick out morphological features. We discuss the discovery of 'Red Hooks' in the pCMDs of six early-type galaxies and two spirals and postulate their origins. We develop quantitative methods to characterize the pCMDs, including measures of the blue-to-red light ratio and colour distributions of each galaxy and we organize these by morphological type. We compare the colours of the pixels in each galaxy with the stellar population models of Bruzual & Charlot to calculate star formation histories for each galaxy type and compare these to the stellar mass within each pixel. Maps of pixel stellar mass and mass-to-light ratio are compared to galaxy images. We apply the pCMD technique to three galaxies in the Hubble Ultra Deep Field to test the usefulness of the analysis at high redshift. We propose that these results can be used as part of a new system of automated classification of galaxies that can be applied at high redshift.     

Objective classification of galaxy spectra using the information bottleneck method

A new method for classification of galaxy spectra is presented, based on a recently introduced information theoretical principle, the information bottleneck . For any desired number of classes, galaxies are classified such that the information content about the spectra is maximally preserved. The result is classes of galaxies with similar spectra, where the similarity is determined via a measure of information. We apply our method to ∼6000 galaxy spectra from the ongoing 2dF redshift survey, and a mock-2dF catalogue produced by a cold dark matter (CDM) based semi-analytic model of galaxy formation. We find a good match between the mean spectra of the classes found in the data and in the models. For the mock catalogue, we find that the classes produced by our algorithm form an intuitively sensible sequence in terms of physical properties such as colour, star formation activity, morphology, and internal velocity dispersion. We also show the correlation of the classes with the projections resulting from a principal component analysis.     

An atlas of predicted exotic gravitational lenses

We present an investigation of the relationships between the radio properties of a giant radio galaxy MRC B0319−454 and the surrounding galaxy distribution with the aim of examining the influence of intergalactic gas and gravity associated with the large-scale structure on the evolution in the radio morphology. Our new radio continuum observations of the radio source, with high surface brightness sensitivity, images the asymmetries in the megaparsec-scale radio structure in total intensity and polarization. We compare these with the three-dimensional galaxy distribution derived from galaxy redshift surveys. Galaxy density gradients are observed along and perpendicular to the radio axis: the large-scale structure is consistent with a model wherein the galaxies trace the ambient intergalactic gas and the evolution of the radio structures are ram-pressure limited by this associated gas. Additionally, we have modelled the off-axis evolution of the south-west radio lobe as deflection of a buoyant jet backflow by a transverse gravitational field: the model is plausible if entrainment is small. The case study presented here is a demonstration that giant radio galaxies may be useful probes of the warm-hot intergalactic medium believed to be associated with moderately over dense galaxy distributions.