How Artificial Intelligence Can Aid in the Search for Dark Matter

Swiss researchers have developed an artificial intelligence algorithm capable of distinguishing signals related to dark matter from confusing background signals derived from astronomical observation data. Created by scientists at the École Polytechnique Fédérale de Lausanne (EPFL), this deep learning algorithm utilizes convolutional neural networks, a powerful type of neural network designed for processing image data.

After training the algorithm on extensive simulated data from a cosmological model, it achieved an impressive 80% accuracy in identifying dark matter signals from galaxy cluster images under ideal conditions. The relevant research has been published in a recent issue of the journal Nature Astronomy. Dark matter is widely recognized as the unseen force that makes up about 85% of all matter in the universe. It neither emits light nor interacts with electromagnetic forces, making direct observation impossible; scientists can only study it indirectly through gravitational effects.

Previous studies have indicated that interactions between dark matter particles could influence their movement and distribution, which can be observed at a galactic scale. Galaxy clusters, which are composed of numerous galaxies, contain vast amounts of concentrated dark matter, making them ideal subjects for dark matter research. However, the multitude of galaxies also introduces considerable "noise." For instance, the energetic emissions from supermassive black holes at the centers of galaxies can affect the motion of surrounding matter, resulting in effects that can easily be mistaken for those caused by dark matter interactions.

This research simulated various scenarios involving dark matter and the feedback effects from active galactic nuclei, the intensely energetic centers of galaxies. By analyzing thousands of simulated galaxy cluster images, the AI algorithm learned to differentiate between signals caused by dark matter interactions and those arising from active galactic nuclei feedback. This breakthrough suggests that artificial intelligence could play a crucial role in analyzing astronomical data, with its adaptability and reliability making it a promising tool for future research on dark matter and other astronomical phenomena.