| Semester | Winter 2022 |
| Course type | Practical Course/ Lab |
| Lecturer | Jun.-Prof. Dr. Wressnegger |
| Audience | Informatik Master & Bachelor |
| Credits | 4 ECTS |
| Time | 14:00–17:15 |
| Room | 149, Building 50.34 and online |
| Language | English |
| Link | https://campus.kit.edu/campus/all/event.asp?gguid=0xBA82196A8149422EBB1381E0C46FAD75 |
| Registration | https://ilias.studium.kit.edu/goto.php?target=crs_1922845&client_id=produktiv |
In this practical course, the students develop learning-based systems for different computer security tasks, thereby intensifying their knowledge gained in the lecture "Machine Learning for Computer Security."
The students have the unique opportunity to design, implement, and evaluate systems based on real-world data used in computer security research.
The "Datalab" is composed of 6 units with several individual tasks
covering different topics from classical computer security research,
such as attack detection, spam classification, or vulnerability
discovery. In each unit, the students develop an approach, train and
validate it on
The best approaches are awarded at the end of the semester and presented at a joint colloquium and get-together.
| Date | Step |
| 24. Oct | Kick-off (physical presence necessary) & Unit 0: An easy start |
| 31. Oct and 7. Nov | Unit 1: Lots and Lots of Spam |
| 14. Nov and 21. Nov | Unit 2: Network-based Attack Detection |
| 28. Nov and 5. Dec | Unit 3: Embedded Malware |
| 12. Dec and 19. Dec | Unit 4: Android Malware Detection |
| 9. Jan and 16. Jan | Unit 5: Adversarial Machine Learning |
| 23. Jan and 30. Jan | Unit 6: Model Stealing/Extraction |
| 6. Feb | Deadline for the last unit |
| 13. Feb | Award ceremony and presentation at final colloquium |
To participate it is strongly recommended to have attended the lecture
or the lecturesNews about the practical course, potential updates to the schedule, and additional material are distributed using a separate mailing list. Moreover, the list enables students to discuss topics and solution approaches.
You can subscribe here.