1: \begin{abstract}

2: Variational inference methods for latent variable statistical models have gained popularity because they are relatively fast, can handle large data sets, and have deterministic convergence guarantees.

3: However, in practice it is unclear whether the fixed point identified by the variational inference algorithm is a local or a global optimum.

4: Here, we propose a method for constructing iterative optimization algorithms for variational inference problems that are guaranteed to converge to the $\epsilon$-global variational lower bound on the log-likelihood.

5: We derive inference algorithms for two variational approximations to a standard Bayesian Gaussian mixture model (BGMM).

6: We present a minimal data set for empirically testing convergence and show that a variational inference algorithm frequently converges to a local optimum while our algorithm always converges to the globally optimal variational lower bound.

7: We characterize the loss incurred by choosing a non-optimal variational approximation distribution suggesting that selection of the approximating variational distribution deserves as much attention as the selection of the original statistical model for a given data set.

8: \end{abstract}

9: