 Stochastic gradient Langevin dynamics

Stochastic Gradient Langevin Dynamics (SGLD) is a method used in the field of machine learning and statistical inference for sampling from a probability distribution, typically a posterior distribution in Bayesian inference. It combines ideas from stochastic gradient descent and Langevin dynamics, which is a form of stochastic differential equations often used in physics to describe the evolution of particles under the influence of both deterministic forces and random fluctuations.