Doctoral student Markus Leyser discusses research question

In a new format, the members of the Research Training Group discuss a concrete research question posed by a member in a plenary session. The aim is to consider this research question from the perspectives of different disciplines and to compare their approaches. Views from other disciplines are explicitly welcome in order to break down existing thought patterns. The question is formulated briefly and concisely in the form of a book title.

Doctoral student Markus Leyser (Bridging project BR3) deals with the modelling of energy markets with the help of machine learning methods. He posed the question of how a catalogue of requirements can be created for this task when the target group itself has no expert knowledge in machine learning. The focus here was on the requirements for the interpretation and communication of the models created. 

One result of the event was a translation of the task into a vocabulary commonly used in engineering sciences. While in the engineering context, for example, one speaks of "specifications", in software projects one more often speaks of "requirements". Potential misunderstandings due to vagueness in the use of such terms can thus be recognized at an early stage. The participants also shared their experiences with methods to address this issue, such as expert interviews. Furthermore, it was discussed in which subject area the results could best be published (machine learning vs. communication science vs. energy industry). Among other things, the extent to which aspects that are difficult to quantify, such as trust in the models, should be taken into account played a role here.

In the course of the day, several discussions also took place between the students on the topics discussed. The Boysen-TU Dresden Research Training Group plans to continue using and actively developing this format in the future.