Data Science for Materials Science
Aims and Scope
Designing data-science methods tailored to materials science or deploying existing data-science methods there will yield a significant scientific benefit, for both sides (computer science and related disciplines as well as materials science). This call for contributions and participation does not list potential topics, to avoid that such a list might look like a selection by the organizers. Instead we want to reach out to and bring together as many KIT scientists as possible. However, workshop contributions should either relate to data science (and not just be computer science or mathematics in general), or materials-science contributions should have a connection to data analytics or be speculative regarding such a connection. -- Having said this, we now illustrate the benefit mentioned earlier by means of some examples:
- Computer-science perspective: There exists data from materials science that computer scientists would refer to as Moving Objects Data (MOD). Think of data describing fractures or deformations whose positions and sizes change over time. A unique feature of such MOD -- in contrast to almost all use cases from other domains -- is that the objects are not points, but regions with a spatial extent. This gives way to a huge research potential which computer scientists could tap.
- Mathematics perspective: Inverse problems in materials sciences naturally arise. Think of the imaging of inner structures with tomographic methods for parameter identification in complex systems based on experiments. Here the objective is the extraction of information for measured data.
- Materials-sciene perspective: In materials science, extensive experiments and simulations abound. 'extensive' means that one repeats the experiment/the simulation with slightly different parameter values many times. Modern data-science methods allow to predict, at least in principle, which experiment/which simulation should take place next -- if one has formalized a research objective. Research on how to adapt the various methods to the various scenarios from materials science will yield a huge benefit.
The workshop envisioned will bring together computer scientists, mathematicians and scientists from other disciplines with an appropriate methodological background who are interested in data science and its deployment in materials sciences as well as materials scientists who have data and are amenable to computer-science methods for its exploitation, i.e., promote networking. The workshop will focus on content. Regarding the formation of formal collaborations in turn, this workshop will only be a first step at best.
- Combination of physical and virtual attendance. The number of individuals who can attend physically will follow from the regulations regarding hygiene which will be valid in January.
- Physical attendance in particular -- again, respecting the regulations regarding hygiene -- shall promote new acquaintances within KIT and personal, rather informal interaction.
- Online participation will be possible if the capacity of the room is exhausted, or if individuals simply prefer to participate in this way. Details regarding the IT infrastructure will follow.
- We are currently evaluating under which conditions catering for individuals who are physically present is feasible.
- Presentations shall last about 15 minutes and be followed by discussions. We seek a program that is as diverse as possible.
- The event is exclusive for KIT members.
- If you want to present, please send email to sekretariat boehm on November 11th 2020 at the latest, including a tentative presentation title. Kindly let us know whether you plan to attend physically or prefer a 'remote' presentation. ∂ ipd kit edu
- If you wish to participate without presenting, please send email to sekretariat boehm on November 30th 2020 at the latest. Please let us know whether you want to attend physically (we will do the allocation in a first-come-first-serve order) or whether a virtual attendance is sufficient for you in any case. ∂ ipd kit edu
Klemens Böhm (klemens boehm) ∂ kit edu
Location and Time: Campus South, Room 145/146 in the AVG building, January 22nd 2021, 9.00 am