Program

Speakers: Timo Koch and Sarbani Roy
Topic: Publishing Simulation Data and Software
Place: V47.02

...Research Software Best Practices

Software is an integral part in much of today's research which naturally poses the question how to publish and archive software as an important part of research results. Often research software is developed by researchers. The software itself may even be the subject of the research. The reproducibility of results obtained with such software (as an integral part of the scientific process) can be more efficiently achieved--researchers never have enough time! --by following a few basic rules during development and publishing. Discussing best practices, I will argue, that following guidelines has the added benefit of aiding the robustness of the research process itself. Moreover, I will try to make the point that for practical reasons, thinking about reusability of developed research software, rather than reproducibility of results, may be a more natural way to create research software that advances the scientific field of interest.  Finally, I will discuss several "modes" of publishing software and how they may create a sustainable development cycle. This discussion includes some experience we gained in a recently completed project in collaboration with an interdisciplinary team of researchers from the University of Stuttgart.

Timo Koch (University of Oslo)

Sarbani Roy (University of Stuttgart/SimTech)

Sarbani Roy is an enthusiast of open science and FAIR research. She holds a doctoral degree in Mathematics from the Indian Institute of Technology Kharagpur (IIT Kharagpur), India. Shortly after, Sarbani joined the Department of Hydromechanics and Modelling of Hydrosystems at the University of Stuttgart as a postdoctoral researcher. Currently, Sarbani holds the position of research data manager at the Stuttgart Center for Simulation Sciences (SC SimTech) and is a scientific employee at the Competence Center for Research Data Management (FoKUS) within the University of Stuttgart.  Her active involvement in the BETTY task area of NFDI4Ing (German National Research Data Infrastructure for engineering science) showcases her commitment to advancing research data management practices. Sarbani possesses a strong research interest in advancing research data management practices, establishing platforms for archiving sustainable research software, and ensuring seamless reproducibility of scientific artifacts. Her keen interest also lies in exploring the role of metadata in ensuring the long-term sustainability of research software. She is deeply committed to advocating for open science principles and empowering researchers to enhance the accessibility, transparency, and reusability of their work. By fostering a culture of openness and facilitating effective research data management, she actively contributes to scientific progress and fosters collaboration. With a strong drive for interdisciplinary collaboration, Sarbani aims to bridge the gap between research and reproducibility.

Speaker: Ulrike von Luxburg
Topic: Explainability and regulations
Place: V47.02

Explainability is one of the concepts that dominate debates about the regulation of machine learning algorithms. In my presentation I will argue that in its current form, post-hoc explanation algorithms are unsuitable to achieve the law's objectives, for rather fundamental reasons. In particular, most situations where explanations are requested are adversarial, meaning that the explanation provider and receiver have opposing interests and incentives, so that the provider might manipulate the explanation for her own ends. I then discuss a theoretical analysis of Shapley Value based explanation algorithms that open the door to more formal guarantees for posthoc explanations.

Ulrike von Luxburg (University of Tübingen)

Ulrike von Luxburg is a full professor for the Theory of Machine Learning at the University of Tuebingen, Germany. Her research analyzes machine learning algorithms from a theoretical point of view, tries to understand their implicit mechanisms, and to give formal statistical guarantees for their performance. In this way, she reveals fundamental assumptions, biases and strenghts and weaknesses of widely used machine learning algorithms, for example in the field of explainable machine learning. Next to her own research group, she is coordinating a large research consortium on Machine Learning in Science. She is an active participant in local debates about ethics and responsibility in machine learning. 

Place: Foyer downstairs

MS 7 "Simulations@Operations - Pervasive simulations: anytime, anywhere"
Place: V 47.03

• Selection optimization for solvers of simulations
  Julia Meißner, Melanie Herschel and Dominik Göddeke, University of Stuttgart
• Interactive Micro-Visualizations with Simulations on Wearable Devices
  Tanja Blascheck, Fairouz Grioui, University of Stuttgart
• Towards the robotic construction site
  Grzegorz Orzechowski, Laparanta LUT University
• Scheduling Pervasive Simulations in Heterogeneous Edge Computing
  Janick Edinger, Universität Hamburg
• Data-driven reduced order modeling
  Nathan Kutz, University of Washington

MS 8 "Accelerating Simulations through Quantum Computing"
Place: V 47.02

• Robustness of quantum algorithms against coherent control errors
  Daniel Fink and Julian Berberich, University of Stuttgart
• Quantum Computational Fluid Dynamics
  Matthias Möller and Merel A. Schalkers, TU Delft
• Accelerating Simulations using Hybrid Quantum-Classical Machine Learning
  Hamzeh Kraus, TWT GmbH Science & Innovation
• Polymer Physics by Quantum Computing
  Cristian Micheletti, Scuola Internazionale Superiore di Studi Avanzati

Place: Pfaffenwaldring 47, Foyer downstairs

Speaker: Miriah Meyer
Topic: Troubling Visualization
Place: V47.02

Visualization research is at an inflection point where the field is filled with increasingly diverse research interests and approaches, but where we are also struggling to make our tried-and-true approaches to research answer an increasingly complex range of questions. How do we consider people’s affective, emotional, and subjective relationships to data and visualization? How do we design novel visualizations in an increasingly complex and uncontrollable technology landscape? What are our ethical responsibilities to our collaborators, our participants, and each other? In this talk I’ll argue that it is time to trouble the foundational perspectives we hold around how we, as researchers, make sense of the world and design within it. I’ll talk about new perspectives we are using in the Vis Collective at Linköping University, and the myriad of research opportunities these perspectives are pointing us to.

Miriah Meyer (Linköping University) 

Speaker: Robert Bitmead
Topic: Particle filters for feedback control - simulation-based tools for state estimation
Place: V47.02

The particle filter is a computational mechanism for propagating empirical probability densities and updating them to incorporate new data using Bayes Rule via resampling. In a stochastic control context, the conditional density of the state is calculated in this way and then used for the construction of the next feedback control value. This latter calculation is in general intractable because it requires solution of the Stochastic Dynamic Programming Equation. So, simplifying and approximate methods are needed. The most dramatic of these is Certainty Equivalence Control, where a specific single “best” value of the state is selected and used without regard to the nature of the rest of the density, such as its variance; often this is chosen to be the conditional mean. Other choices will be presented and described. The complexity of moving from conditional state density to stochastic optimal control also limits the application to popular modern methods such as Model Predictive Control (MPC) because these rely on open-loop constrained optimization. The core issue is that stochastic optimal control is necessarily closed-loop and density based while MPC is an open-loop method in hiding. In the presentation, a middle path is developed which blends open and closed loop while maintaining computational tractability.  Simulation in open- and closed-loop form a part of the procedure to select a best state value. Alongside this, the particle filter operates to provide the correct density for search.

Robert Bitmead (University of California)

Bob Bitmead is Distinguished Professor in Mechanical & Aerospace Engineering at the University of California, San Diego. He holds degrees in Applied Mathematics and Electrical Engineering from Sydney University and Newcastle University, both in Australia. He has held faculty positions at the Australian National University and James Cook University of North Queensland. He is a control theorist with a long experience in control applications in many industrial sectors. His theoretical work is strongly informed and guided by these applications. He was the recipient of the 2014 ASME Rufus Oldenburger Medal and of the 2015 IEEE Controls Systems Transition to Practice Award. Bob was President of the IEEE Control Systems Society for 2019. He was a member of the IFAC Council from 1996 to 2002 and was Foundation Editor-in-Chief of the IFAC Journal of Systems & Control 2017-2022. He is Fellow of IEEE, IFAC and the Australian Academy of Technological Sciences and Engineering. Bob brews his own beer and is an accredited and active Australian Rules Football umpire training with the San Diego Lions Australian Football Club.

Presentation of the Best Poster Award and closing remarks as well as farewell remarks by Miriam Schulte.

Place: Pfaffenwaldring 47
Room: V47.02