INFORMATION FOR STUDENTS OF MACHINE LEARNING

You need to indicate your preferred choice of diploma seminars when recruiting for a second-cycle degree programme in the system of Internet Recruitment of Candidates (IRK). Assignment to the seminar is based on ranking lists.

Registration for classes

You will be registered for the obligatory courses of the winter semester of your first year of study (unless you have previously passed their equivalents) by the Student office. You can choose your course class/lab group during the RDG round (direct registration for groups) or the groups exchange system. The deadlines for the respective registration rounds can be found in our faculty's USOSweb.

From the second semester of study onwards, you will need to register on your own for all courses (except the master seminar). The general section of the student guide contains detailed registration information and course group descriptions.

Bootcamp classes only take place in the first weeks of October, as well as additional hours of DNN classes for students who need more support in this course. This requires some attention when planning your class schedule as all other classes also start from the beginning of October. Look also at the location of each class - this will not always be 2 Banacha Street.

Choice of courses

According to the course schedule valid from the academic year 2025/26, you will need to complete elective courses for 18 ECTS (it was 24 ECTS in previous academic years) (corresponding to 3 standard courses of 6 ECTS each). You can choose these from the elective courses for ML or the elective courses for computer science.

Please note that starting from the 2025/26 academic year, there are new obligatory courses in the first-year programme of ML studies: Professional Skills in Computer Science and Statistical data analysis 2, while Statistical machine learning, Ideas and Informatics and one of the bootcamps have been removed.

Due to a change in the programme of study for the Bachelor of Science in Computer Science, some of the previously obligatory courses will, starting from the academic year 2024/25, become elective courses with new codes: Numerical Methods (1000-2M24MNU), Program Semantics and Verification (1000-2M24SWP), Security of computer systems (now with a new name: Introduction to Cybersecurity and Cryptography - 1000-2M24WCK) and Functional Programming (1000-2M24PFU - which is the equivalent of the former stream in Introduction to fucnctional programming). Students who have taken these courses as obligatory courses during their first cycle studies are asked to choose other elective courses.

In addition, with prior agreement with the vice-dean for student affairs, you can complete a course from Obligatory courses for 1st year 2nd cycle Computer Science, Elective courses for 2nd cycle studies in Mathematics or Elective monographic courses for 2nd cycle studies in Mathematics as an elective course for ML. 

Note that according to the study programme, elective courses are intended to introduce students to advanced techniques with a focus on theoretical computer science or software engineering. This means that having the approval to recognise a course from the groups listed above as an elective course for ML is not automatic. We recommend to make sure about the possibility of recognising a course as an elective one by submitting an application in usos at the beginning of the semester.

If you have completed during your first-cycle studies:

• Deep neural networks with a Course ID: 1000-2M16GSN (ML equivalent: 1000-317bDNN)
• Robot Control with a Course ID: 1000-2M22RC (ML equivalent: 1000-317bRC)
• Visual recognition: neural networks with a Course ID: 1000-2M18RO ( ML equivalent: 1000-318bVR)
• Reinforcement learning 1000-2M20UZW (ML equivalent: 1000-318bRL)
• Natural language processing with a Course ID: 1000-2M21NLP (ML equivalent: 1000-318bNLP)
• Explainable Machine Learning with a Course ID: 1000-1M18WUM (ML equivalent: 1000-319bEML)

you will be exempted from passing the ML equivalent. If you have used the credit to settle a previous study term, you will be exempted from the ML equivalent but you will need to complete another elective course, but if the course has not been linked to a previous study term and not used to settle a previous study term, you can link it yourself to the ML programme.

In addition, in the academic year 2024/25 you may be exempted from passing the SML on the basis of a SAD credited with at least 4; ask the Head of studies (Aleksy Schubert, e-mail: A.Schubert@mimuw.edu.pl) about the rules in the following academic years.

In your second year of studies you should also complete an internship/study visit. The internship supervisor is Dr Waldemar Pałuba (email: W.Paluba@mimuw.edu.pl) and the rules of the completing the internship (and templates of the necessary documents) can be found here, Polish version. Remember, the deadline for completing the internship/study visit is in the winter session.

Diploma examination process

In order to complete the first year of the master seminar, you need to have your master thesis topic approved. In order to complete the 2nd year of the seminar, you need to submit your thesis to the APD (you do not need to print it). You can only take the diploma examination once you have ended studies (i.e. you have completed all the courses required by your study plan). However, you may start arranging the examination date even before you have completed all your credits in Usos.

The master diploma examination is oral. The exam consists of a presentation of the master thesis (up to 15 minutes) and answers to 3 questions directly related to the master thesis topic.