Beskjeder

The exam set and solution are now published further down on the page.

- The only permitted aid on the exam is an approved calculator; you will find the list of approved calculators here: https://www.uio.no/studier/eksamen/hjelpemiddel/mn-math-kalkulatorer.html.

- The formula sheet will be an appendix to the exam paper, and you cannot bring your own copy. You should however take a look at it before the exam; you will find it further down on this web-page.

- Results in the formula sheet can be used without need for being shown, unless you are expressly instructed to do so in the exam problem.

- The exam will consist of sub-problems (a, b, c, ...), that may be of different length and difficulty, but they all count the same on the grade.

- Some of the sub-problems may contain several questions. To get a full score, you must answer all of them.

- A correct answer is not enough in itself to get a fu...

Upon request, I have re-named presentations according to the order in which they were given in the lectures. The presentations are exactly the same as before, just with a different name: STK2100-1, STK2100-2, STK2100-3, STK2100-4, STK2100-5, STK2100-6...

There will be oracle sessions on May 20, 10:15-12:00, NHA room 720 and May 21 12:15-14:00, NHA room 720.

There is no fixed program, but you can come and ask me questions about the curriculum, exam problems or other exercises, or anything else related to the course.

There is an exercise lecture tomorrow (Friday) as usual, and the weekly exercises for this week are now posted.

... is now out (see the link further below on the page). The deadline is Thursday April 23 at 14:30.

Note that the only requirement for you to pass the assignment is that you hand in a report, at the minimum a blank page with your name, course and assignment number. If you want comments to your work, you need to hand in a clearly legible report.

If you need to apply for a postponement of the submission deadline due to illness or other reasons, you have to contact Ingrid Hobæk Haff (e-mail: ingrihaf@math.uio.no), no later than the same day as the deadline at 14:30. Further, use you uio-email to contact me, as your e-mail otherwise risks ending up in the junk folder.

The exercise session tomorrow, March 20, 10:15-12:00, has been moved to room UE32 in Niels Henrik Abels hus. 

I also want to clarify something I said about the backfitting algorithm for additive models during the lecture I held today. The update \(\hat{f}_j \leftarrow \hat{f}_j - \frac{1}{N}\sum_{i=1}^N \hat{f}_j(x_{ij})\) ensures that \(\frac{1}{N} \sum_{i=1}^N \hat{f}_j(x_{ij}) = 0\). By convergence, it is meant that the \(\hat{f}_j\) stabilize. I apologize for any confusion!

Best regards,

Berit

First, note that the book of James et al 2013: An Introduction to Statistical Learning (ISLR), does not cover all the curriculum, nor all that is written in The Elements of Statistical Learning, and is only meant as a supplement, which is potentially easier to read. Below is an approximate correspondence between (parts of) the curriculum and ISLR:

Curriculum:                                 ISLR:

- Chapter 2                                  - Chapter 2

- Chapter 3.1                               - Chapter 3

- Chapter 4                                  - Chapter 4.1-4.4

- Chapter 5                                  - Chapter 7.3-7.5

- Chapter 6                                  - Chapter 7.6

- Chapter 7.10-7.11                     - Chapter 5 

- Chapter 9.1                               - Chapter 7.7

- Chapter 9.2                               - Chapter 8.1

- C...

... is now available further down on the web page. The deadline is March 5 at 14:30.

The lecture onn Wednesday February 11 10:15-12:00 is cancelled, as Auditorium 5 is reserved for another activity. The exercise class on February 13 10:15-12:00 are moved to room UE32, in Niels Henrik Abels hus.

The teaching will each week consist of 4 hours of lectures (Wednesdays and Thursdays), as well as 2 hours of exercise classes on Fridays, where you can get help solving the weekly exercises. Note that there are no exercises in week 4, and hence, no exercise classes on Friday January 23rd.

On the semester page, plans for the lectures will be published in the schedule, and the weekly exercises will be published on "Weekly exercises" under "Exercises".

We will use the text book: Hastie et al (2009): Elements of Statistical learning. The book James et al (2013): An introduction to Statistical learning - with applications in R can be used as supplementary material. The latter is an easier version of the former, with less mathematical details. Both books are available online. Further, the course will be a mix of statistical/mathematical theory and applied work on the computer. We will use R extensively.

Looking forward to seeing you on January 21st!