The course teaches the core techniques for deriving discrete, well-interpretable qualitative models from observed and measured continuous metrics. Qualitative models reflect “engineering thinking” and help to understand the underlying phenomena and causal relationships in a system, identify bottlenecks, etc. As qualitative models are equipped with precise semantics, formal methods are available to reason about them and to establish proofs of correctness.

Computer-based systems are becoming increasingly complex – in the number of their components as well as their interactions. Intelligent algorithms and highly dynamic IT infrastructures further increase complexity. Therefore, ensuring their extra-functional properties during design and operation is fundamental (e.g., efficiency, performability, and dependability). Thus, modeling current systems for design and operation support requires the design- and runtime use of "system identification" techniques in a classic system theoretic context.

The course delivers a theoretical as well as practical overview of identifying qualitative models from observations and measurements; “explaining” models; and reasoning about their correctness. The application of these methods in research as well as in industrial contexts are both covered.

The lectures include hands-on practice sessions for each major topic, based on industrially motivated research challenges.