This workshop provides you with basic strategies and tools for an effective and efficient time and project-management. You will get specific ideas for improving your own approach towards organizing yourself and your work. The motto of the workshop is: Becoming a better, smarter, more focused time-manager is an evolution not a revolution. You already have a lot of the ideas and skills that you need. This workshop helps you to understand them in a deeper way and to commit yourself to really using them.

Contents in brief:
Successful project management (milestone plans vs. iterative incremental approach), basic strategies and tools for an efficient time and self-management (i.e. Pomodoro Technique, phases of productivity, implementation intentions), setting priorities, pragmatism and productivity).

Methods:
Input and discussion, individual and group work, coaching techniques.

Workshop:
Building on familiar classical methods such as linear regression, the workshop will guide participants towards recent ideas in distributional adaptivity and explore their relevance to statistical learning and modern data analysis. Participants with a good understanding of classical statistical methods and an interest in mathematical statistics are warmly encouraged to join.

Keynote "Outrigger Local Polynomial Regression":
Date & Time: 13 October 2026, 16:30
Venue: NTI Lecture Hall, KIT Campus South
The talk will revisit the classical method of local polynomial regression from a modern perspective, showing how it can be adapted to the diverse error distributions encountered in contemporary applications. Along the way, it will introduce the main methodological ideas and explore the mathematical theory underlying the new approach.

In the context of multivariate nonparametric regression with missing covariates, we propose Pattern Embedded Neural Networks (PENNs), which can be applied in conjunction with any existing imputation technique. In addition to a neural network trained on the imputed data, PENNs pass the vectors of observation indicators through a second neural network to provide a compact representation. The outputs are then combined in a third neural network to produce final predictions. Our main theoretical result exploits an assumption that the observation patterns can be partitioned into cells on which the Bayes regression function behaves similarly, and belongs to a compositional Hölder class. It provides a finite-sample excess risk bound that holds for an arbitrary missingness mechanism, and in combination with a complementary minimax lower bound, demonstrates that our PENN estimator attains in typical cases the minimax rate of convergence as if the cells of the partition were known in advance, up to a poly-logarithmic factor in the sample size. Numerical experiments on simulated, semi-synthetic and real data confirm that the PENN estimator consistently improves, often dramatically, on standard neural networks without pattern embedding. Code to reproduce our experiments, as well as a tutorial on how to apply our method, is publicly available.