Challenges and opportunities in high dimensional variational inference

AK Dhaka, A Catalina, M Welandawe… - Advances in …, 2021 - proceedings.neurips.cc
Current black-box variational inference (BBVI) methods require the user to make numerous
design choices–such as the selection of variational objective and approximating family–yet …

Variational inference via Wasserstein gradient flows

M Lambert, S Chewi, F Bach… - Advances in Neural …, 2022 - proceedings.neurips.cc
Abstract Along with Markov chain Monte Carlo (MCMC) methods, variational inference (VI)
has emerged as a central computational approach to large-scale Bayesian inference …

Forward-backward Gaussian variational inference via JKO in the Bures-Wasserstein space

MZ Diao, K Balasubramanian… - … on Machine Learning, 2023 - proceedings.mlr.press
Variational inference (VI) seeks to approximate a target distribution $\pi $ by an element of a
tractable family of distributions. Of key interest in statistics and machine learning is Gaussian …

On the convergence of black-box variational inference

K Kim, J Oh, K Wu, Y Ma… - Advances in Neural …, 2024 - proceedings.neurips.cc
We provide the first convergence guarantee for black-box variational inference (BBVI) with
the reparameterization gradient. While preliminary investigations worked on simplified …

The bayesian learning rule

ME Khan, H Rue - arxiv preprint arxiv:2107.04562, 2021 - arxiv.org
We show that many machine-learning algorithms are specific instances of a single algorithm
called the Bayesian learning rule. The rule, derived from Bayesian principles, yields a wide …

Towards understanding the dynamics of gaussian-stein variational gradient descent

T Liu, P Ghosal… - Advances in Neural …, 2024 - proceedings.neurips.cc
Abstract Stein Variational Gradient Descent (SVGD) is a nonparametric particle-based
deterministic sampling algorithm. Despite its wide usage, understanding the theoretical …

Tractable structured natural-gradient descent using local parameterizations

W Lin, F Nielsen, KM Emtiyaz… - … on Machine Learning, 2021 - proceedings.mlr.press
Natural-gradient descent (NGD) on structured parameter spaces (eg, low-rank covariances)
is computationally challenging due to difficult Fisher-matrix computations. We address this …

Efficient, multimodal, and derivative-free bayesian inference with Fisher–Rao gradient flows

Y Chen, DZ Huang, J Huang, S Reich… - Inverse Problems, 2024 - iopscience.iop.org
In this paper, we study efficient approximate sampling for probability distributions known up
to normalization constants. We specifically focus on a problem class arising in Bayesian …

Weight initialization based‐rectified linear unit activation function to improve the performance of a convolutional neural network model

B Olimov, S Karshiev, E Jang, S Din… - Concurrency and …, 2021 - Wiley Online Library
Abstract Convolutional Neural Networks (CNNs) have made a great impact on attaining
state‐of‐the‐art results in image task classification. Weight initialization is one of the …

Gradient flows for sampling: mean-field models, Gaussian approximations and affine invariance

Y Chen, DZ Huang, J Huang, S Reich… - arxiv preprint arxiv …, 2023 - arxiv.org
Sampling a probability distribution with an unknown normalization constant is a fundamental
problem in computational science and engineering. This task may be cast as an optimization …