Messages
You will find the exam text here and a solution proposal here. We will now start going through and grade the exams. In the meantime, we would be very grateful if you could take a few minutes to fill out this online course evaluation form. Your feedback is very important and it will help us when preparing future versions of the course.
Thank you and hope you all have a nice and relaxing Christmas break! :)
Since there have been a few questions about this, I will add a message also here: you will not be allowed to bring any support material to the exam (which will indeed be a good old-fashioned pen-and-paper exam:). And remember to check in Studentweb where to go for the exam, as it is spread out over multiple locations.
Good luck on Wednesday!
You will find an overview of the material for the exam here. Basically, it includes the lectures slides (except for conformal prediction) and the associated parts in course book, as well as the exercise problems.?Please let me know if you find anything in the list that looks weird/incorrect.
Next week, we will have a final lecture on conformal prediction on Wednesday (26.11) and practice on an old exam on Thursday (27.11).
The second mandatory assignment is now available here. The report should be submitted through Canvas and the deadline for submission is November 6 at 14:30. To pass you need to make an effort at solving the problems, but we will not score your solutions. As before, think about the assignments as additional practical exercises for which you will get direct feedback on your solutions. Please don't hesitate to ask if you have any questions about the problems and/or report (note also that you can update your submitted solutions up until the deadline).
Happy solving!
As a reminder, the deadline for submitting the first mandatory assignment is on Thursday this week (Oct 2nd) at 14:30. Note also that we have no teaching sessions this week, so you may use the time to work on the assignment.
The first mandatory assignment is now available here. The report should be submitted through Canvas and the deadline for submission is October 2 at 14:30. You should think about the assignments as additional exercises for which you will get direct feedback on your solutions. To pass you need to make an effort at solving the problems, but we will not score your solutions. And, if you want to use tools such as ChatGPT, for learning purposes I would strongly recommend using it only as a discussion partner. In particular, you should write the report yourself. Please don't hesitate to ask if you have any questions about the problems and/or report (note also that you can update your submitted solutions up until the deadline).
Happy solving!
In the second lecture week we are going to finish the part on linear models (that we started on the first week) and then move on to model selection. In preparation for the lectures, I would ask you to read (or at least glance through) the relevant sections in the course book, which you find under syllabus in the schedule.
Also, due to the digitalisation lecture (see message below), we will exceptionally start our Wednesday lecture at 10:30, so everyone interested may attend the digitalisation lecture without missing the beginning of our lecture.
The Norwegian Centre for Knowledge-driven Machine Learning Integreat invites you to this year’s Integreat Digitalisation Lecture – an annual tradition where the Minister of Digitalisation and Public Governance presents the state of digitalisation in Norway.
Integreat Digitalisation Lecture 2025: Karianne Tung
Wednesday 3 September 2025, 09:30–10:30
Helga Eng Building, Auditorium 1, Blindern, University of Oslo
UiO Rector Ragnhild Hennum opens the lecture.
Minister Karianne Tung delivers the keynote speech.
Professor Ingrid Glad moderates the panel discussion with the Minister and Integreat researchers.
The event is open to all and requires no registration. Language: Norwegian. Streaming.
More information: Integreat Digitalisation Lecture 2025
Following the first two lectures, you will now have one week to work on the problems in exercise sets 1-2 before the exercise classes next week where you will have a look at some solutions. As I mentioned in class, many of the pen-and-paper problems are quite challenging, so don't be too discouraged if you are not able to fully solve a problem. Already understanding and thinking about the problem is very useful, and then also understanding the solution. Also, due to slow progress by the lecturer, we did not yet cover the lasso, which is part of ex 3.16, so you can consider that as an extra star problem.
Welcome to the 2025 edition of Statistical Learning Methods in Data Science! In this course you will learn about the inner workings of several modern statistical (machine) learning methods, and how to apply these in practice. The course will be based on the book The Elements of Statistical Learning. Starting Aug 20, the plan is to have two lectures every other week and exercise classes every other week. There will be two mandatory assignments that you need to pass to be allowed to taken the final exam, which will be a written exam (pen-and-paper). More info about the course will be given during the first lecture, and feel free to send me (Johan) an email if you have any questions about the course.