Please register for the courses 708.063 (lecture) and 708.064 (exercises) via TUGOnline.


To be announced.

Grading (for the practicals)

There are three assignments (one every two weeks), and a mini-project:

  1. Simulated Annealing, Gradient descent and Genetic-Algorithms
  2. Q-Learning and SARSA
  3. Model based learning and Policy gradient
  4. Mini-Project
The students can make groups of three and will be graded on an oral examination.
Every student of each group have to submit all assignments on canvas .
During the examination the students will present one random assignment and the mini-project.
The grade is based on 50 % on the assignments and 50% on the mini-project.


  1. You must name your team members on the first page of your submitted reports. Each student of each group submit a report written in his/her own words.
  2. Each member of the team must be able to present the solution of every assignment based on his/her own report.
  3. For the mini-project, prepare slides on one of your laptop.
    Your should support your results with working demos, figures or videos and we might ask to see some code.
    Try to have all material tested and ready (5 slides max for your mini-project).
  4. The grade is individual, each student must present a part of the work during the oral presentation. Still the presentation of each member will strongly influence the grade of the rest of the group.


  1. Copying of solutions or parts of solutions will not be tolerated!
  2. The first time a copied solution is found, all involved persons receive zero points for this example. If it happens again, all involved persons will receive 5 negative points for each copied solution.
  3. Code for programming exercises may be shared within teams, as long as everybody writes an own report. Code sharing between teams is not allowed!
  4. If you are not able to present and explain your submitted solution at the examination, you and your team will be penalized accordingly.