Machine Learning techniques in Lattice Gauge Theories

With the advent of GPUs and low-precision computing paradigms,
machine learning methods have become feasible. Computational methods in
lattice gauge theories can be broadly classified as simulation and
measurement methods. In the domain of simulation methods, flow based
methods and autoencoder based methods will be discussed in detail with
several practical drawbacks. In the measurement domain, sign optimized
contour methods and control variate variance reduction techniques will be
reviewed.