The secrets of the Big Bang revealed by artificial intelligence?

The secrets of the Big Bang revealed by artificial intelligence?

It couldn't be more complicated: tiny particles spin wildly with extremely high energy, countless interactions occur in the tangle of quantum particles, and this results in a state of matter known as "quark-gluon plasma". Immediately after the Big Bang, the entire universe was in this state; today it is produced by collisions of high-energy atomic nuclei, for example at CERN.

Such processes can only be studied using high-performance computers and highly complex computer simulations whose results are difficult to evaluate. Therefore, using artificial intelligence or machine learning for this purpose seems like an obvious idea. Normal machine learning algorithms, however, are not suitable for this task. The mathematical properties of particle physics require a very special structure of neural networks. At TU Wien (Vienna), it has now been demonstrated how neural networks can be successfully used for these challenging tasks in particle physics.

“Simulating a quark-gluon plasma in the most realistic way possible requires a extremely large amount of computation time, ”says Dr. Andreas Ipp of the Institute for Theoretical Physics of TU Wien. "Even the largest supercomputers in the world are overwhelmed by this." It would therefore be desirable not to accurately calculate every detail, but to recognize and predict certain properties of the plasma with the help of artificial intelligence.

Therefore, neural networks, similar to those used for image recognition, are used: artificial "neurons" are linked together on the computer in a similar way to neurons in the brain and this creates a network that can recognize, for example, whether a cat is visible in a certain image or not.

When applying this technique to quark-gluon plasma, however, there is a serious problem: the quantum fields used to mathematically describe the particles and the forces between them, can be represented in various different ways. "This is referred to as gauge symmetries," explained Ipp. “The principle behind coò is something we are familiar with: if I gauge a measuring device differently, for example, if I use the Kelvin scale instead of the Celsius scale for my thermometer, I get completely different numbers, even though I am describing the same physical state. It is a bit like with quantum theories, except that the allowed changes there are mathematically much more complicated ". Mathematical objects that appear completely different at first glance can in fact describe the same physical state.

"If these gauge symmetries are not taken into account, it is not possible to interpret the results of computer simulations in a meaningful way" , said Dr. David I. Müller. “Teaching a neural network to understand these gauge symmetries on its own would be extremely difficult. It is much better to start by designing the structure of the neural network in such a way that gauge symmetry is automatically taken into account, so that different representations of the same physical state also produce the same signals in the neural network, ”Müller said. "This is exactly what we have now managed to do: we have developed completely new network levels that automatically take into account the invariance of the meter." In some test applications, it has been shown that these networks can actually learn much better how to handle quark-gluon plasma simulation data.

“With such neural networks, it becomes possible to make predictions about the system, such as example estimate what the quark-gluon plasma will look like at a later time without having to calculate in detail every single intermediate step in time, ”explained Andreas Ipp. "And at the same time, the system is guaranteed to produce only results that do not contradict gauge symmetry in other words, results that make sense at least in principle."

It will take some time before this is possible completely simulate atomic nucleus collisions at CERN with such methods, but the new type of neural networks provides an entirely new and promising tool for describing physical phenomena for which all other computational methods may never be powerful enough.

Powered by Blogger.