The European organization for nuclear research, CERN, recently adopted two new innovations employing artificial intelligence for the enhanced detection and analysis of elementary particles. Both developed by Seth Moortgat at the Vrije Universiteit Brussel (VUB), the first of two methods was designed to make it easier to identify charm quarks, bottom quarks and top quarks – the three heaviest matter particles. The second method was developed to increase the sensitivity of the data analysis when comparing the results and to verify theoretical models faster.
The particle accelerator, known as the Large Hadron Collider (LHC), rapidly speeds up protons, which then collide with each other, simulating conditions like those occurring half a second after the big bang – the starting point for the search into unknown physics phenomena. “The Standard Model of particle physics describes the basic building blocks of the universe and the forces responsible for the interactions between these particles,” explains Moortgat. “However, the model remains incomplete as it fails to describe phenomena such as dark matter, the mass of neutrinos, or even gravity.”
By using the simulations from the LHC, Moortgat conducts fundamental research into the interactions between these phenomena. To show that self-learning machines could recognize specific patterns in large amounts of data, he developed innovative algorithms to identify the different types of quarks in the detector, to distinguish between them and, in some cases, exclude large numbers of new physics models from the measurement results. “With the development of innovative machine learning methods, we can now measure how often the three heaviest quarks are produced together in the billions of particle collisions that occur every second in the LHC,” says Moortgat. Continuing: “Excluding models is extremely important in order to understand which unknown phenomena we apparently still overlook today.”