Software Engineering and Testing for AI Systems

How do you know and assess whether an AI-intensive system is working as intended? On the one hand, quality assurance for such systems can be considered to be a software testing challenge; in parallel, the machine learning components within such systems are normally validated according to machine learning-specific evaluation methodology.

In this course, you will learn more about these two complementary takes to evaluation and quality assurance, while learning about (and getting hands-on experience in) state-of-the-art advances in combining these two takes.

Course content taught includes:

• Basics of testing (assertions, white-box vs. black-box testing)

• Basics of machine learning evaluation (validation en validity)

• Quality assurance at different scopes in AI-intensive systems (models, components, systems, cyber-physical systems)

• Adversarial attacks

• Software testing techniques (e.g. fuzzing, metamorphic testing, combinatorial testing, differential testing)

• Data labels (ground truth/oracle) validation

The final grade of the course consists of the following components:

• Group Report (weighting 70%)

• Digital Exam (weighting 30%)

Final grade calculation = 0.7 * Group Report + 0.3 * Digital Exam

A passing final grade for the course can only be earned when for all components at least a 5.0 is earned, and the weighted final grade is at least a 5.8.

In case of an insufficient final result, repair options may exist in accordance with Article 17A, Times and number of examinations, sub 1, of the Teaching and Examination Regulations, for:

• Group Report: Repair opportunity

• Digital Exam: Resit opportunity

Disclaimer: information may change depending on unforeseen circumstances or measures (see: TER Art 2, sub 5).