Seminar

Towards a machine learning description of nuclei

By
Dr Arnau Rios Huguet (ICC & FQA, UB)
Date
seminar ICCUB
Machine learning techniques are ubiquitous in science, and nuclear physics is not an exception. Neural networks have found a variety of nuclear applications in the past, from capturing systematic trends in data to learning the features of restricted model-space simulations for extrapolation purposes. Recently, machine learning tools have also been used to solve directly for the properties of quantum many-body fermionic and bosonic systems. These developments open the door for a first-principles machine learning description of nuclei. I will discuss recent embryonic applications of these techniques to few-body nuclear systems, and lay down some of the challenges ahead.