Unveiling the Essence of Stochastic in Machine Learning

On the Convergence of the Unadjusted Langevin Algorithm