Messages
The only permitted aid to bring is an approved calculator, which you should bring. Info about approved calculators can be found here (only in Norwegian).
A list of distributions will be given as an Appendix to the exam, along with some useful formulas.
The exam takes place 13 December at 14:30 (4 hours) in Sal C in Silurveien 2.
This will be an open lecture, for example questions from the students. Starts at 13.15, ends before 14.31, depending on the turnout
Comment to Problem 2 a. (iii) With acceptance rates I actually meant acceptance probabilities, which is min(1,r) (with r from (11.2) or (11.1) in the textbook), hence in particular I wanted you to express r for each parameter for the proposal distribution you have chosen. Summaries of the acceptance rates achieved in the simulations should be reported in 2 b. (iv). When correcting exams, confusion about this point will of course be accounted for in a good way
In 2 c. (i) it should of course say ''length tilde(y)'' not "weight tilde(y)". Corrected now.
I have made clear for Problem 1 a. (ii) that you should identify which known distributions all the full conditional distributions equal.
will be available on the course website on Monday 14 November at 14.00, and the solution has to be handed in by Monday 28 November before 14.00. The solutions must be independent work by each candidate, so no cooperation with others.
If you wish to withdraw from the exam you must do so in Studentweb before 14 November. Failure to do so will be counted as one of the three opportunities to sit the exam.
The programming parts of the project exam should be solved using R.
On Thursday 3 November, I will go through the exercises first, and use as much time on that as I feel necessary.
The lecture/exercises session on 13 October is cancelled due to illness
The exam will be a project assignment followed by a written exam. The project assignment will be available on the course webpage Monday 14 November, and the project paper has to be handed in by Monday 28 November. Date and place for the final written exam can be found here.
We have two student representatives: Lena (lenarj "at" student.matnat.uio.no) and Vegard (vegarant "at" math.uio.no)
You need to respond to the Doodle-link I have sent out for when you are going to make your mandatory presentation before 22 September, or the presentation will not be approved by me for you to be admitted to the final exam. If you have not gotten the doodle-link by now, let me know.
I have now sent a doodle-link to those who are currently registered for the stk9021 for planning when each of you are going to present the exercises for that week. You have to respond to this doodle! If you do not get the doodle-link, let me know.
On the coming Thursday, I will again ask one of the PhD students who plans to take the final exam in stk9021 to give a one hour presentation the following Thursday. Because there are so many registered for stk9021, we will do this every week, and the theme will be to go through (some or all of) the exercises for the following week.
On Thursday 8 September we need to find two student representatives for the evaluation of the course. Think about volunteering!
On the coming Thursday, I will ask one of the PhD students who plans to take the final exam in stk9021 to give a one hour lecture (on a theme discussed with me) the following Thursday. This is because of what is stated on the stk9021 course web site: "each phd student is expected to give a one hour oral presentation on a topic of relevance (chosen in cooperation with the lecturer). The presentation has to be approved by the lecturer for the student to be admitted to the final exam." So each PhD student taking the final exam in stk9021 will have to this once during the semester!
The lecture notes from yesterday are now available on the 'Lecture plan' page. I will continue to publish the lecture notes until otherwise announced (I will evaluate how this is working out).
Please do the exercises for next week, which are now announced on the 'Exercises' page.
The lectures will take place Thursdays 12.15-15, starting 25 August.
We will mainly use the textbook "Bayesian Data Analysis" (Gelman, Carlin, Stern, Dunson, Vehtari, Rubin, 3rd edition, 2014), supplied with some additional material during the semester.
Ida Scheel