Upcoming Events
- Thursday, 15.10.2026, 14:00 (WIAS-ESH)
- Seminar Numerische Mathematik
Dr. Jose Rodrigo Rojo Garcia, WIAS Berlin:
Bayesian optimal experimental design with Wasserstein information criteria
more ... Location
Weierstraß-Institut, Anton-Wilhelm-Amo-Str. 39, 10117 Berlin, Erdgeschoss, Erhard-Schmidt-Hörsaal
Abstract
Bayesian optimal experimental design (BOED) is a widely used method for reducing uncertainty in Bayesian inversion. A classical design criterion is the Expected Information Gain, which measures discrepancy between prior and posterior probability measures via Kullback-Leibler divergence. Nevertheless, some limitations arise in this classical approach, for instance, the comparison of probability measures with different supports. In order to mitigate these issues, a new proposal based on the p-Wasserstein distance is proposed as a utility function. The present work establishes results of well-posedness, consistency, interpretation, a closed formula for linear Gaussian case, prior and likelihood stability, empirical measure convergence rate, numerical demonstrations, and some numerical algorithms".
Host
WIAS Berlin
- Wednesday, 21.10.2026, 10:00 (WIAS-ESH)
- Forschungsseminar Mathematische Statistik
Prof. Dr. Kolyan Ray, Imperial College London, GB:
Bayesian regression models with dense confounding variables
more ... Location
Weierstraß-Institut, Anton-Wilhelm-Amo-Str. 39, 10117 Berlin, Erdgeschoss, Erhard-Schmidt-Hörsaal
Abstract
Inferring causal relationships from observational data can be invalidated by the existence of confounding variables. We consider nonparametric Bayesian inference in regression models with additive errors under a dense confounding assumption, i.e. that every confounding variable affects many covariates. We show that by using suitable priors, which are equivalent to marginalizing out the confounders in a certain sense, such confounding can be corrected for. Our approach is illustrated via Gaussian process and spike and slab priors, for which we provide accompanying theoretical guarantees via posterior convergence rates. This is joint work with Luke Travis
Further Informations
Dieser Vortrag findet auch via Zoom statt: https://hu-berlin.zoom-x.de/j/65177102181?pwd=nUMtJOkzDx8xyWzPFezeHh0NEIQUv3.1, Meeting-ID 651 7710 2181, Kenncode: 536200
Host
Humboldt-Universität zu Berlin
Universität Potsdam
WIAS Berlin
- Thursday, 22.10.2026, 10:30 (WIAS-Library)
- Software and Data Seminar
Dr. Laura Prieto Saavedra, WIAS Berlin:
Numerical solution of a Poisson equation using FEniCsx and deal.II
more ... Location
Weierstraß-Institut, Hausvogteiplatz 5-7, 10117 Berlin, R411
Abstract
When developing and studying new numerical methods, it is often useful to have a simple and reliable computational setup for testing ideas, verifying results and comparing discretizations. In this seminar, we will introduce two implementations to solve a Poisson equation numerically using two well-established open-source FEM libraries for solving PDEs: FEniCSx and deal.II. We will go through the math, the basic problem setup, and the essential components of each code, and demonstrate how they can be easily adapted to more difficult problems (e.g., other boundary conditions, other discretizations, etc.). The goal is to provide a simple, reusable starting point that can be modified for different research purposes. No prior FEniCSx or deal.II experience is needed. Basic programming experience is helpful, but not required.
Further Informations
Seminar Software and Data
Host
WIAS Berlin
- Thursday, 22.10.2026, 14:15 (WIAS-ESH)
- Berliner Oberseminar „Nichtlineare partielle Differentialgleichungen” (Langenbach-Seminar)
Clara Thierbach, Humboldt-Universität zu Berlin:
A variational characterization of a Gibbs measure related to the Aviles-Giga functional
more ... Location
Weierstraß-Institut, Anton-Wilhelm-Amo-Str. 39, 10117 Berlin, Erdgeschoss, Erhard-Schmidt-Hörsaal
Host
WIAS Berlin
Humboldt-Universität zu Berlin
Freie Universität Berlin
- Wednesday, 28.10.2026, 11:30 (WIAS-406)
- Seminar Interacting Random Systems
Lars Schroeder, University of Twente:
The Geometry of Hierarchical Navigation
more ... Location
Weierstraß-Institut, Anton-Wilhelm-Amo-Str. 39, 10117 Berlin, 4. Etage, Weierstraß-Hörsaal (Raum: 406)
Abstract
Many applications use multi-layered hierarchical data structures for fast approximate nearest-neighbor search in high-dimensional vector spaces. However, the geometric conditions that ensure accurate and efficient greedy navigation remain poorly understood. In this talk, we study the efficiency of greedy navigation on a hierarchy of proximity graphs constructed from n data points on the d-dimensional torus. We identify a deterministic coverage condition under which, given any query q, greedy search returns a point within (1+eps)-factor of the distance to the closest point. This coverage property holds with high probability when the data is distributed as a homogeneous Poisson process, a Hermitian determinantal process, or a bounded-density Cox process, as long as d=o(logn/loglogn). Additionally, we obtain logarithmic expected hop count for the number of greedy hops in fixed dimension. Joint work with Shankar Bhamidi, Souvik Dhara and Clara Stegehuis.
Host
Humboldt-Universität zu Berlin
Technische Universität Berlin
WIAS Berlin
International Research Training Group 2544
- November 3 – 6, 2026 (WIAS-ESH)
- Workshop/Konferenz: Stochastic processes with reinforcement
more ... Location
Weierstraß-Institut, Anton-Wilhelm-Amo-Str. 39, 10117 Berlin, Erdgeschoss, Erhard-Schmidt-Hörsaal
Host
WIAS Berlin
- Wednesday, 18.11.2026, 11:30 (WIAS-406)
- Seminar Interacting Random Systems
Vinicio Rodrigues Moreira, UFMG Brazil:
tba
more ... Location
Weierstraß-Institut, Anton-Wilhelm-Amo-Str. 39, 10117 Berlin, 4. Etage, Weierstraß-Hörsaal (Raum: 406)
Abstract
tba
Further Informations
Seminar Interactin Random Systems
Host
WIAS Berlin
- Wednesday, 18.11.2026, 14:15 (WIAS-ESH)
- Berliner Oberseminar „Nichtlineare partielle Differentialgleichungen” (Langenbach-Seminar)
Gianna Götzmann, Universität Augsburg:
Perturbed minimizing movements of time-dependent functionals on metric spaces
more ... Location
Weierstraß-Institut, Anton-Wilhelm-Amo-Str. 39, 10117 Berlin, Erdgeschoss, Erhard-Schmidt-Hörsaal
Abstract
Minimizing movement schemes are a fundamental tool for approximating evolution equations that have a gradient-flow structure with respect to an energy. Motivated by homogenization problems involving, for instance, geometries that evolve in time, we consider Gamma-converging perturbations of time-dependent driving energies. We study the convergence of the corresponding minimizing movement scheme and provide sufficient conditions under which limits of the scheme are (time-dependent) curves of maximal slope for the limit energy. We further exhibit examples showing that, in general, such a stability property may fail, with the limit motion depending on the relation between the time-discretization parameter τ and the perturbation parameter ϵ. This is joint work with Jan-Frederik Pietschmann and Antonio Tribuzio.
Further Informations
Oberseminar “Nichtlineare partielle Differentialgleichungen” (Langenbach-Seminar)
Host
Humboldt-Universität zu Berlin
WIAS Berlin
- Wednesday, 25.11.2026, 11:30 (WIAS-406)
- Seminar Interacting Random Systems
Ulrik Thinggaard Hansen, University of Innsbruck:
tba
more ... Location
Weierstraß-Institut, Anton-Wilhelm-Amo-Str. 39, 10117 Berlin, 4. Etage, Weierstraß-Hörsaal (Raum: 406)
Abstract
tba
Host
Humboldt-Universität zu Berlin
Technische Universität Berlin
WIAS Berlin
International Research Training Group 2544