course: Master Practical Course Machine Learning and Security

teaching methods:
practical course
e-learning, Moodle, computer based presentation
responsible person:
Prof. Dr. Thorsten Holz
Prof. Dr. Thorsten Holz (ETIT), M. Sc. Thorsten Eisenhofer (ETIT), M. Sc. Joel Frank (ETIT)
see examination rules
offered in:
winter term and summer term

dates in winter term

  • kick-off meeting: Wednesday the 13.10.2021 from 14:00 to 15.00 o'clock in ID 03/411
  • lab: according to agreement

dates in summer term

  • kick-off meeting: Wednesday the 14.04.2021 from 14:00 to 15.00 o'clock
  • lab: according to agreement


Form of exam:lab
Registration for exam:Directly with the lecturer
continual assessment


The students obtain a profound understanding of modern machine learning techniques and their applications in the area of computer security. More specifically, the participants are proficient in corresponding ML algorithms and can analyze complex problems on their own. The students can design and implement ML algorithms on their own and learn how to perform research in the intersection of machine learning and computer security.


The practical course provides an introduction to various machine learning (ML) techniques and their application in computer security. In six exercises, we plan to cover the following topics:

  • Linear and logistic regression
  • Clustering algorithms (e.g., k-nearest neighbors) and classification algorithms
  • Unsupervised Learning
  • Support vector machines (SVM)
  • Deep Learning
  • Adversarial Machine Learning

We will cover different applications of these techniques in areas such as:

  • Spam classification
  • Malware clustering
  • Deep fake detection

The course will cover tools such as NumPy and PyTorch. We expect that students perform their own research and investigation to solve the exercises.



recommended knowledge

Basic knowledge of Python is strongly recommended. The course Deep Learning offered by Prof. Fischer covers some recommended basics.


There is a mandatory meeting every two weeks during which we present the new exercises. Every other week, we offer an optional meeting to answer questions. All materials for the course are available via Moodle, please register for the course online.

At most 20 students can participate in the practical course. More information on the planned schedule and the formal requirements are discussed in a meeting that takes place in the first week of the semester, please also see the Moodle course.