A Trust-Region Method for Stochastic Variational Inference with Applications to Streaming Data
Abstract
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.
Cite
Text
Theis and Hoffman. "A Trust-Region Method for Stochastic Variational Inference with Applications to Streaming Data." International Conference on Machine Learning, 2015.Markdown
[Theis and Hoffman. "A Trust-Region Method for Stochastic Variational Inference with Applications to Streaming Data." International Conference on Machine Learning, 2015.](https://mlanthology.org/icml/2015/theis2015icml-trustregion/)BibTeX
@inproceedings{theis2015icml-trustregion,
title = {{A Trust-Region Method for Stochastic Variational Inference with Applications to Streaming Data}},
author = {Theis, Lucas and Hoffman, Matt},
booktitle = {International Conference on Machine Learning},
year = {2015},
pages = {2503-2511},
volume = {37},
url = {https://mlanthology.org/icml/2015/theis2015icml-trustregion/}
}