Picture for Giacomo Zanella

Giacomo Zanella

Department of Decision Sciences, Bocconi University, Milan, Italy, Bocconi Institute for Data Science and Analytics, Bocconi University, Milan, Italy

Optimal lower bounds for logistic log-likelihoods

Add code
Oct 14, 2024
Viaarxiv icon

Linear-cost unbiased posterior estimates for crossed effects and matrix factorization models via couplings

Add code
Oct 11, 2024
Viaarxiv icon

Entropy contraction of the Gibbs sampler under log-concavity

Add code
Oct 01, 2024
Viaarxiv icon

Convergence rate of random scan Coordinate Ascent Variational Inference under log-concavity

Add code
Jun 11, 2024
Viaarxiv icon

Robust Approximate Sampling via Stochastic Gradient Barker Dynamics

Add code
May 14, 2024
Viaarxiv icon

Scalability of Metropolis-within-Gibbs schemes for high-dimensional Bayesian models

Add code
Mar 14, 2024
Viaarxiv icon

Partially factorized variational inference for high-dimensional mixed models

Add code
Dec 20, 2023
Viaarxiv icon

Complexity of Gibbs samplers through Bayesian asymptotics

Add code
Apr 14, 2023
Viaarxiv icon

Improving multiple-try Metropolis with local balancing

Add code
Nov 21, 2022
Viaarxiv icon

Robust leave-one-out cross-validation for high-dimensional Bayesian models

Add code
Sep 19, 2022
Figure 1 for Robust leave-one-out cross-validation for high-dimensional Bayesian models
Figure 2 for Robust leave-one-out cross-validation for high-dimensional Bayesian models
Figure 3 for Robust leave-one-out cross-validation for high-dimensional Bayesian models
Figure 4 for Robust leave-one-out cross-validation for high-dimensional Bayesian models
Viaarxiv icon