Write a comment
Dark matter, an unseen force believed to hold the universe together, constitutes approximately 85% of all matter and 27% of the universe's content. Although its gravitational effects on galaxies and cosmic structures are well-documented, the true nature of dark matter remains elusive.
A dominant theory posits that dark matter could be made up of particles that interact only through gravity by Robert Schreiber
Berlin, Germany (SPX) Sep 11, 2024
Dark matter, an unseen force believed to hold the universe together, constitutes approximately 85% of all matter and 27% of the universe's content. Although its gravitational effects on galaxies and cosmic structures are well-documented, the true nature of dark matter remains elusive.
A dominant theory posits that dark matter could be made up of particles that interact only through gravity. However, some scientists believe these particles may occasionally interact with each other in what is known as self-interaction. Detecting such interactions would offer crucial clues about dark matter's properties.
A challenge arises from distinguishing the subtle effects of dark matter self-interactions from those caused by active galactic nuclei (AGN) - the supermassive black holes at the centers of galaxies. AGN feedback can create effects similar to those caused by dark matter, making differentiation difficult.
In a significant step forward, astronomer David Harvey at EPFL's Laboratory of Astrophysics has developed a deep-learning algorithm that can untangle these complex signals. Their AI-based method is designed to differentiate between the effects of dark matter self-interactions and those of AGN feedback by analyzing images of galaxy clusters - vast collections of galaxies bound together by gravity. This innovation is expected to enhance the precision of dark matter research.
Harvey's team trained a Convolutional Neural Network (CNN) - an AI particularly adept at recognizing image patterns - using images from the BAHAMAS-SIDM project, which simulates galaxy clusters under varying dark matter and AGN feedback conditions. By being fed thousands of simulated galaxy cluster images, the CNN learned to distinguish between the signals caused by dark matter self-interactions and those caused by AGN feedback.
The most complex CNN architecture tested, named "Inception," proved to be the most accurate. Trained on two main dark matter scenarios with different levels of self-interaction, it was further validated on models, including a velocity-dependent dark matter scenario.
Inception achieved an impressive accuracy of 80% under ideal conditions, effectively identifying whether galaxy clusters were influenced by self-interacting dark matter or AGN feedback. Even with the introduction of observational noise, which simulates future telescope data, the AI maintained high performance.
The success of Inception, and the AI approach overall, demonstrates potential for analyzing the vast amounts of data generated by space observations. Its adaptability and reliability make it a promising tool for future dark matter studies.
AI-based approaches like Inception could significantly impact our understanding of what dark matter actually is. As new telescopes continue to provide unprecedented data, these methods may help scientists quickly and accurately sift through it, potentially unveiling dark matter's true nature.
Research Report:A deep-learning algorithm to disentangle self-interacting dark matter and AGN feedback models
Related Links
EPFL Laboratory of Astrophysics
Stellar Chemistry, The Universe And All Within It