Gaussian Mixture Models (GMMs) are a way to model an empirical distribution of data with a mixture of Gaussians.