inf5402 - Trustworthy Machine Learning (Complete module description)

Module label	Trustworthy Machine Learning
Modulkürzel	inf5402
Credit points	6.0 KP
Workload	180 h
Institute directory	Department of Health Services Research
Verwendbarkeit des Moduls	Master Data Science and Machine Learning (Master) > Kernbereich Wahlpflichtmodule Master's Programme Business Informatics (Master) > Akzentsetzungsmodule der Informatik Master's Programme Computing Science (Master) > Angewandte Informatik
Zuständige Personen	Strodthoff, Nils (module responsibility) Lehrenden, Die im Modul (Prüfungsberechtigt)
Prerequisites	Content requirements are basic theoretical knowledge in machine learning, practical programming knowledge in Python and basic knowledge in training deep neural networks.
Skills to be acquired in this module	Professional competence The students have an overview of the various aspects that determine the quality of machine learning algorithms. are familiar with methods to measure different quality aspects and, if necessary, methods to enhance them, and they can implement and apply these methods. Methodological competence The students independently develop theoretical and practical concepts with the help of in-person events, provided materials, and specialized literature. Social competence The students can present solution approaches for problems in this area to the plenary and defend them in discussions. Self-competence The students are able to assess their own subject-specific and methodological competence. They take responsibility for their competence development and learning progress and reflect on these independently. In addition, they independently work on learning content and can critically reflect on the content. General compentence goals ++ knowledge of data science/ML methods and its foundations ++ Ability to analyze problems, compare and select solution methods + formalizing problems mathematically, developing and implementing solutions, interpret their results + knowledge of ethical, legal, security-related limitations + data presentation & discussion + Scientific literature (reading & writing) + scientific communication skills (in particular with people outside the field of study)
Module contents	Machine learning algorithms are increasingly being applied in a wide range of areas, particularly in safety-critical domains. However, the quality of these algorithms is rarely systematically examined. The focus of this event is on various quality dimensions for machine learning algorithms, especially deep neural networks. This ranges from performance measurement to interpretability/explainability (XAI), robustness (adversarial robustness, non-adversarial robustness, distribution shifts, OOD-detection), uncertainty quantification, fairness/bias, and privacy. The methods will be introduced theoretically in the lecture and practically implemented and applied in the exercises.
Literaturempfehlungen	As there is no single textbook that covers all topics in this lecture series, relevant specialized readings will be recommended throughout the course.
Links	/
Language of instruction	English
Duration (semesters)	1 Semester
Module frequency	Summer term
Module capacity	unrestricted
Teaching/Learning method	V+Ü

Lehrveranstaltungsform	SWS	Frequency	Workload of compulsory attendance
Lecture	2	SoSe	28
Exercises	2	SoSe	28
Präsenzzeit Modul insgesamt			56 h

Examination	Prüfungszeiten	Type of examination
Final exam of module	At the end of the lecture term	Written or oral Exam Active participation: Handing in exercises