Variational Inference with Numerical Derivatives: variance reduction through coupling
This work provides an incremental improvement for researchers and practitioners in machine learning by enabling more efficient variational inference for specific non-Gaussian distributions like Wishart, Gamma, and Student.
The paper tackles the problem of high gradient variance in variational inference for non-Gaussian exponential families by introducing VIND, a method that extends the reparameterization trick using numerical derivatives and coupling, resulting in reduced variance and improved posterior approximations as confirmed by experiments.
The Black Box Variational Inference (Ranganath et al. (2014)) algorithm provides a universal method for Variational Inference, but taking advantage of special properties of the approximation family or of the target can improve the convergence speed significantly. For example, if the approximation family is a transformation family, such as a Gaussian, then switching to the reparameterization gradient (Kingma and Welling (2014)) often yields a major reduction in gradient variance. Ultimately, reducing the variance can reduce the computational cost and yield better approximations. We present a new method to extend the reparameterization trick to more general exponential families including the Wishart, Gamma, and Student distributions. Variational Inference with Numerical Derivatives (VIND) approximates the gradient with numerical derivatives and reduces its variance using a tight coupling of the approximation family. The resulting algorithm is simple to implement and can profit from widely known couplings. Our experiments confirm that VIND effectively decreases the gradient variance and therefore improves the posterior approximation in relevant cases. It thus provides an efficient yet simple Variational Inference method for computing non-Gaussian approximations.