arXiv research

New scalable MCMC sampling for nonsymmetric DPPs speeds up computations.

This paper introduces a neural sampler for scalable sampling from complex distributions.

Scalable method completes ill-conditioned matrices from few samples.

Scalable TS using sparse GPs improves efficiency without sacrificing performance.

A scalable algorithm for sampling and fine-tuning models using Tilt Matching.

This paper introduces a scalable benchmark for evaluating local posterior sampling in neural networks.

This work tackles scalable sampling for nonsymmetric DPPs.

The practical usage of reinforcement learning agents is often bottlenecked by the duration of training time. To accelerate training, practitioners often turn to distributed reinforcement learning architectures to parallelize and accelerate the training process. However, modern methods for scalable reinforcement learnin…

A new method improves Bayesian deep learning by balancing scalability and accuracy.

Improving scalability and stability of Stein discrepancies for scalable goodness-of-fit testing

PolytopeWalk library efficiently samples high-dimensional polytopes.

Novel method for scalable neural network-based blackbox optimization.

DIGing-SGLD improves SGLD for scalable Bayesian learning in dynamic networks.

Conventional sampling techniques fall short of drawing descriptive sketches of the data when the data is grossly corrupted as such corruptions break the low rank structure required for them to perform satisfactorily. In this paper, we present new sampling algorithms which can locate the informative columns in presence …

New graph representation learning network improves scalability and feature integration.

The Graph Convolutional Network (GCN) model and its variants are powerful graph embedding tools for facilitating classification and clustering on graphs. However, a major challenge is to reduce the complexity of layered GCNs and make them parallelizable and scalable on very large graphs -- state-of the art techniques a…

There has been recent interest in developing scalable Bayesian sampling methods such as stochastic gradient MCMC (SG-MCMC) and Stein variational gradient descent (SVGD) for big-data analysis. A standard SG-MCMC algorithm simulates samples from a discrete-time Markov chain to approximate a target distribution, thus samp…

New method improves sampling from noisy energy models.

New method efficiently samples Gaussian process posteriors without cubic scaling.

New coin sampling method for Bayesian inference without learning rates.

SLIM efficiently solves overidentified models in a scalable manner.

A new method for assessing Bayesian sampling quality, PSD, is proposed and shown to be more powerful and efficient.

Paper proposes scalable algorithm to estimate intervention targets in linear models.

We propose a new scalable algorithm for holistic linear regression building on Bertsimas & King (2016). Specifically, we develop new theory to model significance and multicollinearity as lazy constraints rather than checking the conditions iteratively. The resulting algorithm scales with the number of samples $n$ in th…

Thompson sampling (TS) is a class of algorithms for sequential decision-making, which requires maintaining a posterior distribution over a model. However, calculating exact posterior distributions is intractable for all but the simplest models. Consequently, efficient computation of an approximate posterior distributio…

Improved sampling for high-dimensional posteriors with underdamped Langevin.

SwISS improves scalability of Bayesian inference for large datasets.

Unified framework for scalable black-box optimization.

Improved DDPMs achieve high log-likelihoods and sample quality with fewer passes.

ESAC combines genetic methods with RL to improve scalability and efficiency.

New algorithm optimizes interventions under network interference, scaling to large networks.

Increasingly complex datasets pose a number of challenges for Bayesian inference. Conventional posterior sampling based on Markov chain Monte Carlo can be too computationally intensive, is serial in nature and mixes poorly between posterior modes. Further, all models are misspecified, which brings into question the val…

Ensemble++ uses shared-factor ensembles to scale Thompson Sampling for linear and nonlinear bandits.

Bullseye Polytope improves clean-label poisoning attacks in transfer learning.

Improves BED scalability for implicit models.

Scalable algorithm for sampling Gaussian processes using sparse grids and preconditioners.

Develops scalable methods to assess sensitivity and uncertainty in continuous treatment effects.

DTS improves inference-time alignment of diffusion models with less compute.

NIPA aims to translate brain learning mechanisms into scalable Bayesian inference.

New method samples DPPs efficiently without downsampling or low-rank approximations.

Scalable Bayesian sampling is playing an important role in modern machine learning, especially in the fast-developed unsupervised-(deep)-learning models. While tremendous progresses have been achieved via scalable Bayesian sampling such as stochastic gradient MCMC (SG-MCMC) and Stein variational gradient descent (SVGD)…

Meta-learning interpretable decision trees with synthetic data.

Bayesian inference engines improve density estimation accuracy and scalability.

Improved Thompson Sampling algorithms for bandits with tighter regret bounds.

SCE improves network embedding using sparsest cut for negative samples only.

Develops coresets for scalable multivariate distribution estimation.

This thesis tackles non-convex Bayesian learning via scalable dynamic importance sampling algorithms.

Graphs are ubiquitous real-world data structures, and generative models that approximate distributions over graphs and derive new samples from them have significant importance. Among the known challenges in graph generation tasks, scalability handling of large graphs and datasets is one of the most important for practi…