Efficient solutions of fermionic systems using artificial neural networks

In this study, we explore the similarities and differences between variational Monte Carlo techniques that employ conventional artificial neural network representations of ground-state wave function for fermionic systems. Our primary focus is on shallow architectures, specifically restricted Boltzmann machine, examine unsupervised learning algorithms are appropriate modeling complex many-body correlations. We assess advantages drawbacks functions by applying them to a range circular quantum dot findings, which include results systems containing up 90 electrons, emphasize efficient implementation these methods both homogeneous heterogeneous high-performance computing facilities.