194.101 Machine Learning Algorithms and Applications
This course is in all assigned curricula part of the STEOP.
This course is in at least 1 assigned curriculum part of the STEOP.

2024S, PR, 2.0h, 3.0EC


  • Semester hours: 2.0
  • Credits: 3.0
  • Type: PR Project
  • Format: Hybrid

Learning outcomes

After successful completion of the course, students are able to...

  • understand and summarise research papers,
  • derive the needed information to (re)implement learning algorithms,
  • develop implementations/applications of learning algorithms,
  • apply them to data sets and/or in applications,
  • empirically evaluate (and experiment with) machine learning algorithms,
  • identify appropriate hyperparameters for the algorithms, and
  • compare different learning algorithms to analyse their strengths and weaknesses

...if they choose an applied project.

After successful completion of the course, students are able to...

  • summarise and present theoretical properties of a machine learning algorithm,
  • identify theoretical weak points of a learning algorithm,
  • independently study and solve specific theoretical problems,
  • apply theoretical results, and
  • check assumptions made by the algorithms

...if they choose a theoretical project.

Subject of course

The students can choose between a theoretical or applied project (or a combination).

You can find some suggestions for projects on our homepage. We also look forward to hearing your own creative and concrete project ideas (check criteria on homepage).

The goal of the applied project is to understand, (re)implement, and apply machine learning algorithms. Selected algorithms should be evaluated with a variety of hyperparameters, data sets, and/or applications. Expected results could include:

  • comparisons of different algorithms,
  • design of benchmarks, or
  • application of algorithms to (creative) use cases.

The goal of the theoretical project is to work on specific theoretical research questions in the area of machine learning. Expected results could include:

  • formal guarantees for a certain learning algorithm (sample, query or computational complexity bounds),
  • worst-case instances where the algorithm in question provably performs badly, or
  • formalisation of underlying assumptions of an algorithm.

Teaching methods

The main part of this project consists of implementing existing learning algorithms or work on theoretical results. During the project, the students will continuously give progress presentations and receive formative feedback. Finally, the students will submit a report summing up the outcome and present their results in form of a poster session.

Mode of examination


Additional information

3ects -> 75h
8h literature search and proposal writing
16h preparing and attending presentations and poster session, and project meetings
38h work on the project
13h writing the project report



Course dates

Fri13:00 - 15:0008.03.2024 Seminar room (FB0210), 2nd floor, Erzh.Johann-Platz 1Kickoff meeting
Fri13:00 - 15:0012.04.2024 Seminar room (FB0210), 2nd floor, Erzh.Johann-Platz 1Introductory presentations
Fri13:00 - 15:0017.05.2024 Seminar room (FB0210), 2nd floor, Erzh.Johann-Platz 1Progress presentations
Thu13:00 - 15:0013.06.2024 Seminar room (FB0210), 2nd floor, Erzh.Johann-Platz 1Poster session

Examination modalities

The final grade is derived from the quality of

  • the implementation (runnability, scalability, runtime, documentation) or the developed theoretical results,
  • the presentations,
  • the poster session, and
  • the submitted written report.

Course registration

Begin End Deregistration end
19.02.2024 23:59 15.03.2024 23:59 31.03.2024 23:59

Registration modalities




No lecture notes are available.

Previous knowledge