To content

Program of TRR 391 Conference 2026

We have a three-day scientific program in a linear plenary format, i.e., there are no parallel sessions. Please find the full program below.

9:30amWelcome and Coffee
10:00am

Session on High Dimensions and Regularization (TC02)
Chair: Christoph Hanck

10:00am–10:40am
Invited Speaker: Ines Wilms
Cross-temporal forecast reconciliation using machine learning

10:40am–11:00am
Monika Zimmermann
Scalable estimation of large generalized additive models: Extended Fellner-Schall via Hutch++ and conjugate gradients

11:00am–11:20am
Benedikt Lütke Schwienhorst
Diffusion-based denoising beats vanilla score matching in parameter estimation: A theoretical explanation

11:20am–11:40am
Axel Bücher
Dimension reduction in multivariate extremes with an application to wind energy production

11:40am–12:00pm
Ignacio Moreira-Lara
A structural matrix autoregressions framework for international spillovers

12:00pmLunch
1:00pmKeynote: Robert Metcalfe
Title TBA
2:00pmCoffee Break
2:30pm

Session on Prediction (TC06)
Chair: Florian Ziel

2:30pm–3:15pm
Invited Speaker: Andreas Pick
Title TBA

3:15pm–3:40pm
Timo Puschmann
Monitoring distributional forecasts

3:40pm–4:05pm
Christian Schulz
Online regularized generalized linear models for (vine) copulas

4:05pm–4:30pm
Yannick Hoga
A robust test for equal predictive accuracy

5:30pm

Welcome Reception

We invite you to get together at the beach bar Stadthafen nearby the conference venue for a BBQ and drinks (free of charge until 10:00pm).

9:00am

Session on Experimental Design (TC01)
Chair: Kirsten Schorning

9:00am–10:00am
Invited Speaker: Vasiliki Koutra
Designing experiments on networks

10:00am–10:25am
Justin Lamberti
Optimal designs for large time-dependent networks

10:25am–10:50am
Robin Solinus
Optimal design for stochastic processes

11:00amCoffee Break
11:30amKeynote: Stefan Nickel
Towards data driven location science (DDLS)
12:30pmLunch
1:30pm

Session on Monitoring (TC04)
Chair: Matei Demetrescu

1:30pm–2:10pm
Invited Speaker: Wolfgang Schmid
Monitoring matrix-valued data

2:10pm–2:20pm
Marvin Napps
Monitoring Kalman filter-based state estimation in power distribution grids

2:20pm–2:30pm
Souhir Ben Amor
Regularized distributional recurrent neural network with linear skip connection model for electricity price forecasting

2:30pm–2:40pm
Bastian Amberg
Towards closing the visibility gap: Leveraging flight operations data for adaptive logistics

2:40pm–2:50pm
Sheila Görz
From time series to spatial data: A generalized approach for change detection

2:50pm–3:10pm
Fabian Schmidt
Real-time monitoring for stock return predictability in nonstationary volatility environments

3:10pm–3:30pm
Vasyl Golosnoy
Online monitoring of multivariate realized volatility models

3:30pmCoffee Break
3:45pm

Spotlight and Discussion on Open Problems in Energy and Transport
Chair: TBA

3:45pm–4:05pm
Invited Speaker: Layla Martin
Title TBA

4:05pm–4:25pm
Open Topics in 'Transport' - with reference to electric vehicles

4:25pm–4:45pm
Open Topics in 'Energy' - with reference to logistics networks/supply needs

4:45pm

Poster-Session with Beer and Fingerfood (no full meal)

Johannes Bao
Parameter optimization for Kalman filter based distribution system state estimation using expectation-maximization algorithm

Pauline Baur
Optimal design of experiments for computer experiments in logistics using GAMLSS

Sultan Mahmud Chomon
Networked spatial effects in European electricity price forecasting

Paul Elhorst
Assessing the parameter estimates of linear-in-means and spatial econometric models

Lara Meyn-Grünhagen
Bridging events and covariates: An interpretable transformer model for multivariate Hawkes processes

Jacob Grytzka
A GAMLSS approach for interpretable emulation of climate-vegetation relationships

Eva Hümmecke
How much compensation do households demand for giving up price stability? Risk premia in electricity plan choice

Marco Jeschke and Leia Betting
Probabilistic modeling and forecasting of household energy consumption

Lara Kuhlmann de Canaviri
Bayesian optimization for decision support in transport logistics guided by global sensitivity analysis

Steffen Maletz
glmSTARMA - An R package for fitting spatio-temporal models based on generalized linear models

Alexander Marx
Contrastive learning recovers causal features for instrumental variable regression

Lennard Maßmann
A super learner for heterogeneous net average treatment effects

Fabian Mies
Doubly-multiscale nonparametric changepoint inference

Maximilian Mowe
Experimental design for the simulation of logistical terminals

Nils Weitzel
Inferring the spatio-temporal structure of climate variability

Christian Weiß
INGARCH models for spatio-temporal counts or ordinal data

The poster session will end around 6:30pm.