Advancements in artificial intelligence (AI) have opened new horizons for space exploration, particularly in the domain of astrometry. This research investigates the integration of AI techniques, specifically deep neural networks, with space astrometry using the Cassini-Huygens images database. The primary objective is to establish a robust algorithm for the detection and classification of...
The supernova remnant SN1006 has been studied extensively by various X-ray instruments and telescopes due to its historical record, its proximity, and its brightness. In order to accurately study the properties of this remnant itself, it is essential to obtain a detailed and denoised view of its small-scale structures, given the existing observations. Here, we present a Bayesian...
Accurately determining the mass distribution within galaxies is crucial for understanding their formation and evolution. Previous research has traditionally relied on analytical equations based on the Jeans equation to estimate the enclosed mass with minimum projection effect. In this study, we present a novel approach to predict the enclosed mass within a given radius using a machine learning...
Millions of serendipitous X-ray sources have been discovered by modern X-ray observatories like Chandra, XMM-Newton, and recently eROSITA. For the vast majority of Galactic X-ray sources the nature is unknown. We have developed a multiwavelength machine-learning (ML) classification pipeline (MUWCLASS) that uses the random forest algorithm to quickly perform classifications of a large number of...
Current and future ground-based cosmological surveys, such as the Dark Energy Survey (DES), and the Vera Rubin Observatory Legacy Survey of Space and Time (LSST), are predicted to discover thousands to tens of thousands of strong gravitational lenses. The large number of strong lenses discoverable in future surveys will make strong lensing a highly competitive and complementary cosmic probe....
During the process of star formation, a wide variety of molecules can form. The use of ALMA interferometer has made it possible to detect a richness of complex organic molecules (COMs) towards hot cores and hot corinos by studying their rotational transitions. However, the analysis of such spectra is a tedious work and actual technics are not optimal, especially for analyzing a large sample of...
The ΛCDM model stands as the prevailing framework in cosmology, yet discrepancies between Cosmic Microwave Background (CMB) and late universe probes underscore incomplete understanding of essential cosmological parameters, like Ωm and σ8, which govern matter density and density fluctuations in the Universe. To address the limitations of traditional statistical methods, we have developed a...
We present CosmoPower-JAX, a JAX-based implementation of the CosmoPower framework, which accelerates cosmological inference by building neural emulators of cosmological power spectra. We show how, using the automatic differentiation, batch evaluation and just-in-time compilation features of JAX, and running the inference pipeline on graphics processing units (GPUs), parameter estimation can be...
Recent serendipitous discoveries in X-ray astronomy such as extragalactic fast X-ray transients, Quasi-periodic eruptions, extroplanetary transits, and other rare short-duration phenomena in the X-ray sky highlight the importance of a systematic search for such events in X-ray archives. Variable-length time series data in form of X-ray eventfiles present a challenge for the identification of...
Supermassive black holes reside in the center of almost every galaxy. Today's supermassive black holes are mostly dormant (like the one at the center of our Milky Way), but in the past, they were actively accreting large amounts of matter and releasing vast amounts of energy. Galaxies with the brightest, most active supermassive black holes, called active galactic nuclei (AGN), are the most...
Galaxy edges/truncations are Low Surface Brightness (LSB) features located in the galaxy outskirts that delimit the distance up to where the gas density enabled efficient star formation. Therefore, they constitute true galaxy edges. As such, they could be interpreted as a non-arbitrary means to determine the galaxy size, and this is also reinforced by the smaller scatter in the galaxy...
Galaxy groups are gravitationally bound structures composed of galaxies and a hot X-ray-emitting gas that envelops the entire group. These systems are balanced with gravitational potential pulling inwards and thermal pressure from the hot gas pushing outwards. Questions remain about how this balance is altered when galaxies within the group undergo periods of star formation or when...
Upcoming surveys are predicted to discover galaxy-scale strong lenses on the magnitude of 10$^5$, making deep learning methods necessary in lensing data analysis. Currently, there is insufficient real lensing data to train deep learning algorithms, but training only on simulated data results in poor performance on real data. Domain adaptation can bridge the gap between simulated and real...
Next generation instruments are focused on producing massive amounts of spectroscopic data that require new approaches that are computationally efficient and more accurate. While traditional processes such as the convolution-based template matching have been proven successful, they are computationally demanding. Machine learning methods have proven to be orders of magnitude faster and showing...
We study the connection between the factors regulating star formation in galaxies on different spatial and temporal scales and connect morphological features (such as bars, bulges and spiral arms) with their integrated star formation on different timescales. This is being done using machine learning methods, specifically using convolutional neural networks (CNNs). The network is trained on a...
The 3D distribution of galaxies encodes key cosmological information that can probe the growth and expansion history of the Universe. In my talk, I will present how we can leverage simulations and machine learning to go beyond current analyses and extract the full cosmological information of the next-generation galaxy surveys. In particular, I will present SimBIG, a forward modeling framework...
Elaborate simulations of physical systems can be approximated by deep learning model emulators, aka surrogate models, based on training data generated from the full model. Because of powerful deep learning libraries and the enormous speed-up to compute model components or the full likelihood, model emulators becoming more common in astronomy. An interesting computational property of deep...
The BAO feature is damped by non-linear structure formation, which reduces the precision with which we can infer the BAO scale from standard galaxy clustering analysis methods. A variety of techniques, known as BAO reconstruction, have been proposed to mitigate this damping effect; however, in order to work, these methods need to make assumptions abut bias and cosmology as well as to rely on...
Galaxy clusters are a powerful probe of cosmological models. Next generation large-scale optical and infrared surveys will reach unprecedented depths over large areas and require highly complete and pure cluster catalogs, with a well defined selection function. We have developed a new cluster detection algorithm YOLO-CL, which is a modified version of the state-of-the-art object detection deep...
We present a technique to improve the accuracy and training efficiency of normalizing flows for multiple images in the context of cosmology. Normalizing flows are powerful deep generative models that can learn complex probability distributions through invertible transformations applied to a simple distribution. They are well-suited for both image generation and density estimation, enabling...
Large, freely available, well-maintained data sets have made astronomy a popular playground for machine learning projects. Nevertheless, robust insights gained into both machine learning and physics could be improved by clarity in problem definition and establishing workflows that critically verify, characterize and calibrate machine learning models. We provide a collection of guidelines for...
Stellar disk truncations are a long-sought galactic size indicator based on the radial location of the gas density threshold for star formation, i.e., the edge/limit of the luminous matter in a galaxy. The study of galaxy sizes is crucial for understanding the physical processes that shape galaxy evolution across cosmic time. Current and future ultradeep and large-area imaging surveys, such as...
The detection of exoplanets has become one of the most active fields in astrophysics. Despite the fact that most of these discoveries have been made possible through indirect detection techniques, the direct imaging of exoplanets using 10-meter-class ground-based telescopes is now a reality. Achieving this milestone is the result of significant advances in the field of high-contrast imaging...
Finding low-scatter relationships in properties of astrophysical systems is important to estimate their masses/distances. I will show how interpretable ML tools like symbolic regression can be used to expeditiously search for these low-scatter relations in abstract high-dimensional astrophysical datasets. I will present new scaling relations between properties of galaxy clusters that we...
How much cosmological information does a cube of dark matter contain? Are we utilising the full potential of information available within a density field? Neural summaries aim to extract all these informations; but success depends on the availability of simulations, network architecture and hyperparameters, and the ability to train the networks. Even for the simplest summary statistics power...
The damping wing signature of high-redshift quasars in the intergalactic medium (IGM) provides a unique way of probing the history of reionization. Next-generation surveys will collect a multitude of spectra that call for powerful statistical methods to constrain the underlying astrophysical parameters such as the global IGM neutral fraction as tightly as possible. Inferring these parameters...
The study of exoplanet atmospheres plays a vital role in understanding their composition. However, extracting accurate atmospheric parameters from transmission spectra poses significant challenges. Bayesian sampling algorithms, although effective, can be time-consuming and laborious. As an alternative, machine learning techniques offer promising avenues to expedite and enhance this process. ...
The line emission mapper (LEM) is a proposed X-ray probe for high spectral resolution survey observations targeting galaxies and clusters of galaxies to characterise the circumgalactic and intergalactic medium better. The mission will use a microcalorimeter array with 1-2 eV resolution, capturing individual emission lines and offering the ability to spatially map elemental emission within...
Observations have established intriguing correlations between supermassive black holes (SMBHs) and their host galaxies. However, state-of-the-art cosmological simulations have revealed discrepancies in the slope, amplitude, and scatter of the scaling relations when compared to both observational data and among different simulations. Understanding the underlying physical mechanisms responsible...
The rise of cutting-edge telescopes such as JWST, the Large Synoptic Survey Telescope (LSST), and the Nancy Grace Roman telescope (NGRT) has introduced a new era of complexity in the realm of planning and conducting observational cosmology campaigns.
Astronomical observatories have traditionally relied on manual planning of observations, e.g., human-run and human-evaluated simulations for...
Machine learning, and in particular deep neural networks (DNNs), have become primary tools for automatic annotation and analysis of astronomical data. Given that astronomy have been becoming increasingly more dependent on Earth-based and space-based digital sky surveys generating vast pipelines of astronomical data, a large number of DNN-based solutions have already been proposed and applied....
The relationship between galaxies and halos is central to describing galaxy formation and a fundamental step toward extracting precise cosmological information from galaxy maps. However, this connection involves several complex processes that are interconnected. Machine learning methods are flexible tools that can learn complex correlations between a large number of features but are...
For a deep understanding of the Universe, it is crucial to rely on complete and accurate information on its primary constituents. These constituents, such as galaxies, black holes, supernovae, and other compact objects, show distinct features in the sky and therefore imprint differently on astronomical data. In this work, we leverage these differences to construct statistical models for their...
The Euclid space telescope will measure the shapes and redshifts of billions of galaxies, probing the growth of cosmic structures with an unprecedented precision. However, the increased quality of these data also means a significant increase in the number of nuisance parameters, making the cosmological inference a very challenging task. In this talk, I discuss the first application of Marginal...
