Machine Learning to Detect New Natural Phenomena
Machine Learning to Detect New Natural Phenomena
Belgian physicist Seth Moortgat has developed two methods using machine learning, or artificial intelligence, to analyze data generated by CERN's particle accelerator and detect new natural phenomena.
Even today, scientists continue to build on Marie Curie’s legacy in their own fields of research, sharing her passion for science and progress. This is the case for Belgian physicist Seth Moortgat, an experimental particle physicist at CERN, the European Organization for Nuclear Research in Geneva. He has developed two methods using machine learning, or artificial intelligence, to analyze data generated by CERN’s particle accelerator and detect new natural phenomena.
Insignificant and Important at the same time
Seth Moortgat is a postdoctoral researcher at VUB and also works as an experimental particle physicist at CERN, the European Organization for Nuclear Research based in Geneva. There, scientists from around the world study the universe to better understand what it is made of and how it works.
“To this day, we still do not know what 95% of the universe is made of. Being involved in research that seeks to answer these questions makes me feel both insignificant and important at the same time.”
In search of the foundations of the Universe
Particle accelerators are used to study the fundamental building blocks of the universe. The world's largest particle accelerator is located at CERN and is known as the Large Hadron Collider (LHC). With a circumference of 27 kilometres, this underground ring allows scientists to study the results of proton collisions.
Protons are subatomic particles. Together with neutrons, they form the nucleus of an atom. Protons themselves are made up of even smaller particles called quarks: up quarks and down quarks.
Inside the particle accelerator, scientists increase the energy of protons until they reach speeds approaching that of light (300,000 km/s), and then allow them to collide. During these high-energy collisions, tiny particles are released. Detectors positioned at the collision points, functioning like digital cameras, capture images of the collisions and the resulting debris.
Each collision generates an enormous amount of data, and its scientific analysis helps researchers gain a deeper understanding of the smallest components of the universe. This approach has led to historic discoveries, such as the Higgs boson in 2012, a particle believed to be responsible for giving mass to all matter.
Detecting Natural Phenomena more efficiently
Thanks to Seth Moortgat’s work, all the data generated by the particle accelerator can be analyzed even more efficiently. The two methods he has developed make it possible to detect patterns more quickly, including those that could not previously be predicted.
- The first method focuses on identifying different types of quarks. Using an intelligent algorithm, these elementary particles are detected and distinguished from one another after a collision. Scientists can then measure how frequently each type of quark is produced during collisions and study the underlying natural phenomena.
- The second method compares the results of different collisions. This makes data analysis more precise and enables theoretical models to be confirmed or ruled out more quickly. Excluding models is particularly valuable because it helps identify unexplained or previously overlooked natural phenomena.
Thanks to this innovative application of machine learning, natural phenomena can be detected more effectively.
“Subatomic physics is thriving, and so are the technological developments associated with it. I am convinced that the future will bring many exciting advances, including applications that can benefit society. I am very optimistic and excited about what lies ahead,” says Seth Moortgat.
“Subatomic physics is thriving, and so are the technological developments associated with it. I am convinced that the future will bring many exciting advances.”