BayesFlow: Learning complex stochastic models with invertible neural networks

ST Radev, UK Mertens, A Voss… - IEEE transactions on …, 2020 - ieeexplore.ieee.org
Estimating the parameters of mathematical models is a common problem in almost all
branches of science. However, this problem can prove notably difficult when processes and …

Projected Stein variational gradient descent

P Chen, O Ghattas - Advances in Neural Information …, 2020 - proceedings.neurips.cc
The curse of dimensionality is a longstanding challenge in Bayesian inference in high
dimensions. In this work, we propose a {projected Stein variational gradient …

HINT: Hierarchical invertible neural transport for density estimation and Bayesian inference

J Kruse, G Detommaso, U Köthe… - Proceedings of the AAAI …, 2021 - ojs.aaai.org
Many recent invertible neural architectures are based on coupling block designs where
variables are divided in two subsets which serve as inputs of an easily invertible (usually …

Enabling uncertainty quantification in a standard full-waveform inversion method using normalizing flows

C Sun, A Malcolm, R Kumar, W Mao - Geophysics, 2024 - library.seg.org
To maximize the utility of seismic imaging and inversion results, we need to compute not
only a final image but also quantify the uncertainty in the image. Although the most thorough …

Generalized self-concordant analysis of Frank–Wolfe algorithms

P Dvurechensky, K Safin, S Shtern… - Mathematical Programming, 2023 - Springer
Projection-free optimization via different variants of the Frank–Wolfe method has become
one of the cornerstones of large scale optimization for machine learning and computational …

Low-rank tensor reconstruction of concentrated densities with application to Bayesian inversion

M Eigel, R Gruhlke, M Marschall - Statistics and Computing, 2022 - Springer
This paper presents a novel method for the accurate functional approximation of possibly
highly concentrated probability densities. It is based on the combination of several modern …

ASPIRE: Iterative Amortized Posterior Inference for Bayesian Inverse Problems

R Orozco, A Siahkoohi, M Louboutin… - arXiv preprint arXiv …, 2024 - arxiv.org
Due to their uncertainty quantification, Bayesian solutions to inverse problems are the
framework of choice in applications that are risk averse. These benefits come at the cost of …

[PDF][PDF] Uncertainty quantification of material imperfections: surrogates, upscaling and inference

R Gruhlke - 2023 - depositonce.tu-berlin.de
I like to thank all the people that supported me during my studies and my doctoral research.
First and foremost I wish to show my gratitude to Martin Eigel for his mentoring and support …

[PDF][PDF] Explicit and adaptive Bayesian inversion in hierarchical tensor

MRW Marschall - depositonce.tu-berlin.de
This thesis is concerned with the expressibility of high dimensional densities arising in
Bayesian inverse problems by employing low-rank tensor techniques and controlling the …

[PDF][PDF] Stein Variational Newton & other Sampling-Based Inference Methods

R Scheichl - ricam.oeaw.ac.at
The (physical) model gives π (y| x), the conditional probability of observing y given x.
However, to predict, control, optimise or quantify uncertainty, the interest is often really in π …