On a class of Gibbs sampling over networks

B Yuan, J Fan, J Liang, A Wibisono… - The Thirty Sixth …, 2023 - proceedings.mlr.press
We consider the sampling problem from a composite distribution whose potential (negative
log density) is $\sum_ {i= 1}^ n f_i (x_i)+\sum_ {j= 1}^ m g_j (y_j)+\sum_ {i= 1}^ n\sum_ {j …

DP-Fast MH: Private, fast, and accurate Metropolis-Hastings for large-scale Bayesian inference

W Zhang, R Zhang - International Conference on Machine …, 2023 - proceedings.mlr.press
Bayesian inference provides a principled framework for learning from complex data and
reasoning under uncertainty. It has been widely applied in machine learning tasks such as …

Asymptotically optimal exact minibatch metropolis-hastings

R Zhang, AF Cooper, CM De Sa - Advances in Neural …, 2020 - proceedings.neurips.cc
Metropolis-Hastings (MH) is a commonly-used MCMC algorithm, but it can be intractable on
large datasets due to requiring computations over the whole dataset. In this paper, we …

Advances in approximate inference: combining VI and MCMC and improving on Stein discrepancy

W Gong - 2022 - repository.cam.ac.uk
In the modern world, machine learning, including deep learning, has become an
indispensable part of many intelligent systems, hel** people automate the decision …

Where is the normative proof? Assumptions and contradictions in ML fairness research

AF Cooper - arxiv preprint arxiv:2010.10407, 2020 - arxiv.org
Across machine learning (ML) sub-disciplines researchers make mathematical assumptions
to facilitate proof-writing. While such assumptions are necessary for providing mathematical …

Improving sampling accuracy of stochastic gradient MCMC methods via non-uniform subsampling of gradients

R Li, X Wang, H Zha, M Tao - arxiv preprint arxiv:2002.08949, 2020 - arxiv.org
Many Markov Chain Monte Carlo (MCMC) methods leverage gradient information of the
potential function of target distribution to explore sample space efficiently. However …

Skip the Steps: Data-Free Consistency Distillation for Diffusion-based Samplers

PJ Dube, A Bera, R Zhang - openreview.net
Sampling from probability distributions is a fundamental task in machine learning and
statistics. However, most existing algorithms require numerous iterative steps to transform a …