This material supports teaching efforts at the master level for future AI System Engineers to build safe and certifiable systems. It covers basics on probability theory, probabilistic model-building, simulation, robust modules, uncertainty propagation, automatic differentation, deep probabilistic frameworks as well as privacy measures and basics for affective AI. It complements existing courses on machine learning and system and software engineering.
A pedestrian detection example is used to demonstrates the workflow for robust change detection and quantitiave model-based performance measures. Especially the expected performance of a system and the assumptions that are made are highlighted.
