Exercises

Weekly exercises. Exercises with only numbers refer to exercises from the textbook.

Earlier exams are given here.

The book webpage provide solutions to some of the exercises here

Solutions to some exercises are given in the files below. These will be continuously updated when new exercises are given

Solution to text book exercises

Solutions to exercises from note of Nils Lid Hjort

Solutions to exam exercises

 

Exercises for Monday November 6

  • Chapter 14: Exercises 1, 3
  • Exam STK4021 2021: Exercise 1
  • Exam STK4021 2018: Exercise 2
  • Exam STK4021 2017: Exercise 3
  • Exam STK4021 2013: Exercises 2, 3
  • We use most of the time on both Monday and Wednesday to discuss exercises

Exercises for Monday October 31:

  • Chapter 7: 4, 5 (both of these exercises do give quite length calculations, do the others first)
  • Chapter 6: 1, 2 (do exercise 2.13 first), 3
  • Exam STK4021 2021: Exercise 3, 4
  • Challenge: 6.7

Exercises for Monday October 17:

Exercises for Monday October 10:

Exercises for Monday October 3:

  • Chapter 5: 12, 13. Training on calculation within hierarchical models
  • Chapter 10: 2, 3. For understanding various aspects of Monte Carlo
    • In 10.2, you may use that the variance of a q quantile is q*(1-q)/(S*f(z_q))^2) where f(z_q) is the density in the q-quantile z_q.
  • Exercise 12 from 'Course Notes and Exercises by Nils Lid Hjort'. More on understanding loss functions and decision theory
    • I was a bit fast on part c here. The posterior distribution becomes somewhat more difficult in this case. See solutions for a detailed derivation of this. I will also mention this on the Wednesday lecture.
  • Exercise 15 from 'Course Notes and Exercises by Nils Lid Hjort'. For understanding why rejection sampling works. Less central than the other exercises. 
    • Note: The original solutions for 15.f was wrong. This is corrected now.
  • notebook

Exercises for Monday September 26:

  • Chapter 4: 15
  • Exam STK4021 Fall 2018: Exercises 1 and 3
    • In 1f the risk function is the expected loss where the expectation is taken with respect to p(y|theta). However, the calculations here are a bit lengthy and can be skipped.
  • Chapter 5: 3, 8, 11

Exercises for Monday September 19:

  • Chapter 3: 3, 7, 9, 13 (Note: something strange with exercise 7 here)
  • Exam STK4021 Fall 2020: Problem 1, 2
    • I was made aware of that there was an error in Problem 1b (which also has some small consequences for 1c). An updated solution is available here.
  • Chapter 4: 1, 2

Exercises for Monday September 12:

  • Exercises from chapter 2: 10,  11, 12, 
  • Exercises from chapter 3: 1, 9, 15

Exercises for Monday September 5:

  • Exercises from chapter 2: 1, 3, 5, 7, 8, 14
    • R-scripts, see here

Exercises for Monday August 29:

  • Exercises from chapter 1: 1, 2, 3, 6, 7

 

    Published Aug. 2, 2022 1:26 PM - Last modified Nov. 7, 2022 11:35 AM