A trust-region method for stochastic variational inference with applications to streaming data
This addresses a bottleneck in Bayesian inference for practitioners dealing with complex models and streaming data, but it is incremental as it builds on existing stochastic variational inference methods.
The paper tackled the problem of stochastic variational inference being prone to local optima and sensitivity to hyperparameters by replacing the natural gradient step with a trust-region update, resulting in generally better results and reduced sensitivity, and applied it to streaming data where the method was crucial for good performance.
Stochastic variational inference allows for fast posterior inference in complex Bayesian models. However, the algorithm is prone to local optima which can make the quality of the posterior approximation sensitive to the choice of hyperparameters and initialization. We address this problem by replacing the natural gradient step of stochastic varitional inference with a trust-region update. We show that this leads to generally better results and reduced sensitivity to hyperparameters. We also describe a new strategy for variational inference on streaming data and show that here our trust-region method is crucial for getting good performance.