A team of astronomers from the University of Hawaii at the Monoa Institute for Astronomy (IFA) has created a very comprehensive astronomical imaging catalog of stars, galaxies and quarks. Artificial Intelligent Neural Network.
Group University of Hawaii Astronomers at the Monoa Institute of Astronomy (JOFA) Published a List Has 3 billion space objects in 2016, Including stars, galaxies and quarks (active cores of supermassive black holes). Needless to say, analyzing this huge database filled with 2 data petabytes was an inappropriate task for punitive humans and even grade students. One of the main goals of the 2016 catalog release was to better reflect these distant specs of light, and also to map the configuration of the galaxy in all three dimensions. The Pan-Stars team can now check out these items from their to-do list, due to the power of machine learning. The results of their work have come Published To the monthly instructions of the Royal Astronomical Society.
His PS1 telescope located on the summit of Helikali on the Big Island of Hawaii, Capable of scanning 75% of the sky, and is currently conducting the world’s largest deep multicolor optical survey, according to a Press announcement Put out by the University of Hawaii. In contrast, the Sloan Digital Sky Survey (SDSS) covers only 25% of the sky.
To provide the computer with a frame of reference and to teach how to distinguish the spatial class of objects from one another, the team turned to publicly available spectroscopic criteria. Leading author of the study and former IFA’s cosmology postdoctoral fellow, Robert Beck, explained in a press release that these measures of the size and shape of a number of objects in the millions are explained in the press release.
G / O media can get commission
“Using advanced optimization optimization algorithms, we took advantage of a spectroscopic training set of about 4 million light sources to teach neural networks to predict source types and galaxy distances, while at the same time improving for light fading through dust. The Milky Way, ” Beck said.
These training sessions worked well; The sorting of objects led to a neural network being assigned a success rate of .1.1.1% for galaxies, 97.7% for stars, and 96.6.6% for quarks. The system also determined the distances to the galaxies, which were only closed by about 3%. According to the University of Hawaii, the resulting work is “the world’s largest three-dimensional astronomical imaging list of stars, galaxies and quarks.”
Kenneth Chambers, a member of the team and a researcher, explains, “This beautiful map of the universe provides an example of how Pan-Stars can increase the power of large data sets with artificial intelligence techniques and complementary observations. “As Pan-Stars collects more and more data, we will use machine-learning to get more information about objects close to Earth, our solar system, our galaxy and our universe.”
The new list, made possible by a grant from the National Science Foundation, is available to the public Mikulsky Archive for Space Telescope. The database is 300 gigabytes in size, and can be accessed in a number of formats, including downloadable computer-readable tables.
T.His survey has already yielded interesting findings, including an explanation for the spooky area of space known as the cold spot. Using the PS1 Telescope, and also using NASA’s Wide Field Survey Explorer Satellite, Pan-Stars scientists have discovered a giant superwide – a “vast area 1.8 billion light years ahead, in which the density of the galaxy is much lower than normal in the known universe,” as the University of Hawaii Described That was five years ago. According to the researchers, this is the superwide that is causing the cold spot, as it is seen in the cosmic microwave background.
The updated map will also be used to study the overall geometry of the universe, to further examine our theories about the standard cosmological model, and to analyze ancient galaxies, among many other avenues of astronomical and cosmological research.
.
