AI removes most of the noise from astrophysical data

Astronomers in Japan have used a supercomputer and an AI model to remove noise from the data. This will allow you to study galaxies in detail.

Japanese astronomers have developed a new artificial intelligence (AI) method that removes noise in astronomical data caused by random variations in the shape of galaxies. After extensive training and testing on large mock data created with a supercomputer, they applied the new tool to real data from Japan’s Subaru telescope.

They found that the mass distribution obtained using this method is consistent with those accepted in the models of the universe. A powerful new tool will be used to analyze big data from current and planned astronomical surveys.

The data from large-scale surveys can be used to study the large-scale structure of the Universe by measuring the laws of gravitational lensing. With gravitational lensing, the gravity of a foreground object, such as a galaxy cluster, can distort the image of the background object.

But this method of studying images of many galaxies runs into a problem: some galaxies just look strange. Therefore, it is difficult to distinguish an image of a galaxy distorted by gravitational lensing from a galaxy that is indeed distorted. Scientists call this shape noise, one of the major limiting factors in research examining the large-scale structure of the universe.

To compensate for the noise, a team of Japanese astronomers used ATERUI II, the world’s most powerful supercomputer. They then added realistic noise and trained artificial intelligence to statistically reconstruct lensing dark matter based on simulated data.

After training, the AI ​​was able to recover previously unnoticed fine details, which helped improve understanding of cosmic dark matter. Then, using this AI on real data covering 21 square degrees of the sky, the team found a foreground mass distribution consistent with the standard cosmological model.

If you have found a spelling error, please, notify us by selecting that text and pressing Ctrl+Enter.

Leave a Comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Author: John Kessler
Graduated From the Massachusetts Institute of Technology. Previously, worked in various little-known media. Currently is an expert, editor and developer of Free News.
Function: Director
John Kessler

Spelling error report

The following text will be sent to our editors:

36 number 0.271365 time