activity
20242026
collaborators

5 papers

cs.LG2026

How to Approximate Inference with Subtractive Mixture Models

Lena Zellinger, Nicola Branchini, Lennert De Smet +3

Classical mixture models (MMs) are widely used tractable proposals for approximate inference settings such as variational inference (VI) and importance sampling (IS). Recently, mix…

cs.LG2026

Multi-Marginal Flow Matching with Adversarially Learnt Interpolants

Oskar Kviman, Kirill Tamogashev, Nicola Branchini +3

Learning the dynamics of a process given sampled observations at several time points is an important but difficult task in many scientific applications. When no ground-truth trajec…

stat.CO2025

Towards Adaptive Self-Normalized Importance Samplers

Nicola Branchini, Víctor Elvira

The self-normalized importance sampling (SNIS) estimator is a Monte Carlo estimator widely used to approximate expectations in statistical signal processing and machine learning. T…

cs.LG2025

Scalable Expectation Estimation with Subtractive Mixture Models

Lena Zellinger, Nicola Branchini, Víctor Elvira +1

Many Monte Carlo (MC) and importance sampling (IS) methods use mixture models (MMs) for their simplicity and ability to capture multimodal distributions. Recently, subtractive mixt…

stat.CO2024

Generalizing self-normalized importance sampling with couplings

Nicola Branchini, Víctor Elvira

An essential problem in statistics and machine learning is the estimation of expectations involving PDFs with intractable normalizing constants. The self-normalized importance samp…