Trading Off Non-Functional Properties of Machine Learning

The aim of the project is to establish an interdisciplinary research centre at the Johannes Gutenberg University Mainz. Here, interactions and dependencies of different properties of machine learning are to be analysed and weighed. Decisions made by algorithms will be examined with regard to transparency and fairness as well as data protection and the efficient use of resources such as electricity. The focus is on competing needs, for example: How decentralised can data be stored and processed to protect privacy, to what extent does this affect energy consumption? Various conflicting goals are identified and characterised in order to create workable compromises for the application. Ethical and legal aspects are to be considered. The solutions found are to be put to use in an AI lab at Mainz University of Applied Sciences.