Seminars in 2021
Bayesian inference in machine learning
Vladimir Spokoiny, WIAS Berlin online10:00 - 12:00
Statistical inference for binary data is one of the central problem in machine learning which can be treated within a nonparametric Bernoulli model.…
more ›Global sensitivity analysis in environmental modeling
Kim Aleksandra, ETH Zürich and Paul Scherrer Institute, Switzerland online10:15 -11:15
In today’s complex world and economy, international supply chains of goods cause global environmental impacts. Life Cycle Assessment (LCA) is a well…
more ›Analysis of stochastic gradient descent in continuous time
Jonas Latz, University of Cambridge, UK online10:15 - 11:15
Stochastic gradient descent is an optimisation method that combines classical gradient descent with random subsampling within the target functional.…
more ›Stability of UQ and Bayesian inverse problems
Bjoern Sprungk, TU Bergakademie Freiberg online10:15 - 11:15
For partial differential equations with random coefficients we investigate the sensitivity of the distribution of the random solution with respect to…
more ›Theoretical perspectives on actin waves
Arik Yochelis, Ben-Gurion University of the Negev, Israel 2.28.0.10810:15 - 11:45
tba
invited by Carsten Beta
more ›On recently developed non-Gaussian priors and sampling methods with application to industrial tomography
Lassi Roininen, Technische Universität Lappeenranta, Finland 2.14.0.2110:15 - 11:15
We consider two sets of new priors for Bayesian inversion and machine learning: The first one is based on mixture of experts models with Gaussian…
more ›Gaussian likelihoods for ’intractable’ situations
Heikki Haario, Technische Universität Lappeenranta, Finland 2.28.0.108 (Campus Golm, building 28, room 0.108)10:00 - 11:00
Various modelling situations – including chaotic dynamics, stochastic differential equations, random patterns such as produced by the Turing…
more ›Some statistical inference results for interacting particle models in a mean-field limit
Marc Hoffmann, Université Paris-Dauphine, France hybrid event, in person at WIAS (please register) and online; please contact Andrea Fiebig (fiebig[at]math.hu-berlin.de) for details10:00 - 12:00
We propose a systematic | theoretical | statistical analysis for systems of interacting
diffusions, possibly with common noise and/or degenerate…
A modified discrepancy principle to attain optimal rates under white noise
Tim Jahn, Universität Bonn, Germany HU, Johann von Neumann-Haus (Rudower Chaussee 25), room: 3.00813:15 - 14:45
We consider a linear ill-posed equation in the Hilbert space setting under white noise. Known convergence results for the discrepancy principle are…
more ›Parameter Estimation for Semilinear Stochastic Partial Differential Equations
Gregor Pasemann, Humboldt-Universität zu Berlin online via zoom, please contact Prof. S. Reich for the link10:15 - 11:15
A theory of parametric inference for semilinear stochastic partial differential equations is developed, with special emphasis put on diffusivity…
more ›A duality formulation for filter stability
Jin Won Kim, University of Illinois at Urbana-Champaign, USA online via Zoom, pls contact Wilhelm Stannat if you would like to join16:15
In this talk, I revisit the problem of filter stability based on certain dual optimal control-type reformulations of the nonlinear filtering problem.
…
Hypothesis testing on high-dimensional spheres: the Le Cam approach
Davy Paindaveine, Univerité libres de Bruxelles, Belgium hybrid, online- please contact Andrea Fiebig (HU) for the zoom link10:00 - 12:00
Hypothesis testing in high dimensions has been a most active research topics in the last
decade. Both theoretical and practical considerations make it…