Modalities
Lectures
This course has 13 lectures which have an irregular schedule. They do not run for the entire semester, and not every lecture slot is used. Typically, they are Mondays and Tuesdays at 8:30 in E2.2 Günter-Hotz-Hörsaal. Check the timetable for the ground truth.
The first lecture takes place in the second semester week on Monday, April 17th, 2023. The slides used in the lecture will be accessible online.
Attendance in the lectures is not strictly mandatory, but recommended. We will not provide lecture recordings, but the lecture notes will be made available in the CMS.
Exercises
Starting with the day of the first lecture, a new exercise sheet is released every Monday.
These exercises are seperated into two parts: assignments and training exercises. The assignments are mandatory exercises which you need to bring and show competence in during the colloquia. The training exercises are voluntary exercises targeted at deepening your understanding of the course's content.
Colloquia
Instead of submitting solutions for the exercise sheets, groups of students need to partake in colloquia where they will discuss their solutions with the tutors. On the Wednesday nine days after the release of an exercise sheet, the tutors will host a colloquium for each group. The colloquia are 15-minute interviews in which your tutor will quiz you on the topics covered by the corresponding exercise sheet. Every member of the group needs to be appropriately prepared.
You will form groups of three to four students to partake in a colloquium. From April 17th 10:00 to April 20th 23:59 you can enter your group and time slot preferences on your Personal Status page in the CMS. Smaller groups, or students without group preferences, will be merged with other groups. Soon after, you will receive your assigned colloquium slot. We will try our best to respect your group and timeslot preferences.
You pass the colloquium by convincing your tutor that you are keeping up with the content of the course. Passing 11 of the total 12 colloquia is necessary to be admitted to the exam. Every student of a group is graded individually.
Project
Throughout the semester, we will release a data science project. Passing the project is necessary to be admitted to the exam.
Outstanding project submissions may be eligible for bonus exam points. More information t.b.a.
Exam and Exam Admission
The main exam will take place on the 01.08.2023 from 14:00 s.t. The exam will be physically on campus. At a future time, you will be assigned seating in the CMS.
To be admitted to the exam, you must pass at least 11/12 colloquia and get a passing grade on the practical project.
Here are some details on the exam:
- Content / Difficulty:
- You have 90 minutes and up to 90 points to score. Approximately, you have to work one minute for one point.
- All eight chapters are relevant to the exam to the same extent. The tasks might look like they are focused on the later chapters, but obviously, the earlier ones give the foundation for the later ones.
- A (minor) part of the exam could be multiple choice.
- There will be 1 to 3 chilis next to a task, indicating how advanced the task is (more = harder). You could use this information to determine the order in which you approach the tasks. They could also help you figure out why you get stuck on a task and whether you should postpone it.
- Your programming skills (in particular using specific APIs) have been tested in the colloquia assignments. We do not expect you to write perfect Python code in the exam and know specific APIs by heart. Instead, we expect that you can write (pseudocode) programs that illustrate how you would implement a specific algorithm (e.g. given X_i, return the sample variance S^2). At the same time, you are generally not allowed to use the convenient functions, e.g.
numpy.mean(except if the task tells you to use them).
- Tools:
- You are not allowed to use a pocket calculator.
- You are allowed to use a handwritten, double-sided DIN A4 cheat sheet. This sheet can contain anything, including statistics jokes (e.g. "Statistics show that those who celebrate more birthdays live longer.").
- We provide additional exam pages that contain the following:
- CDFs for standardized normal distribution, t distribution, chi^2 distribution.
- Logarithm laws
- Summary Tables for Statistical Tests (Z, t, chi^2 )
Office Hours and Tutorials
Instead of weekly tutorials where a tutor discusses the exercises, we offer office hours where you can get answer to (most of) your questions. Solutions to the exercises will be published after the respective colloquia.
You can find our office hours in the timetable.