TMA4267 Linear statistical models Spring 2025

Welcome to the course TMA4267 Linear statistical models. See General information for contact information, curriculum and teaching materials. The page Exercises and assignments will be updated regularly with both recommended exercises and compulsory assignments.

The first lecture will be given on Tuesday January 7th from 14:15 - 16:00 in S1.

Note that the first exercise session is Thursday the 16th.

Lecture 1 (07.01.2025): Introduction to multivariate distributions and some moments (HS 4.1 and 4.2, see also HS 2.1-2.3)
Lecture notes 07.01.25.
Lecture 2 (09.01.2025): Multivariate moments (HS 4.2) and Mahalanobis transformation (HS 3.3, also HS 2.1-2.3)
Lecture notes 09.01.25.
Lecture 3 (14.01.2025): Mahalanobis transformation (HS 3.3) and Principal components (HS 11.1). Lecture notes 14.01.25 Repetition week 2.
Lecture 4 (16.01.2025): Principal components (HS 11.1-11.3) and Transformations (HS 4.3). Lecture notes 16.01.25. Note on principal components.
Lecture 5 (21.01.2025): Transformations (HS 4.3). Moment generating functions (HS 4.2). Multivariate normal distribution (HS 4.4, 5.1).Lecture notes 21.01.25. Repetition week 3.
Lecture 6 (23.01.2025): Multivariate normal distribution (HS 4.4, 5.1). Lecture notes 23.01.25.
Lecture 7 (28.01.2025): Multivariate normal distribution (HS 4.4, 5.1). Lecture notes 28.01.25. Repetition week 4.
Lecture 8 (30.01.2025): Multivariate normal distribution (HS 4.4, 5.1). Estimation and assessment (HS 3.3, 4.5). Idempotent matrices. Lecture notes 30.01.25.
Lecture 9 (04.02.2025): Quadratic forms and idempotent matrices (FKLM Appendix B, Th. B2, B8). Multiple linear regression,introduction (FKLM 3.1). Lecture notes 04.02.25. Repetition week 5.
Lecture 10 (06.02.2025): Multiple linear regression,introduction and estimation (FKLM 3.1, 3.2). Lecture notes 06.02.25.
Lecture 11 (11.02.2025): Estimation of parameters, properties, useful quadratic forms (FKLM 3.2). Notes on regression.
Lecture notes 11.02.25. Repetition week 6.
Report reference group meeting 1.
Lecture 12 (13.02.2025): Distribution of sum of squares, F distribution and F tests (FKLM 3.2, 3.3). Lecture notes 13.02.25.
Lecture 13 (18.02.2025) ANOVA decomposition, F-test for regression coefficients, Partial F-test, t-test about coefficients.(FKLM 3.2, 3.3, 3.5). Lecture notes 18.02.25. Repetition week 7.
Lecture 14 (20.02.2025) Variable selection criteria. (FKLM 3.2, 3.3, 3.5). Lecture notes 20.02.25.
Lecture 15 (25.02.2025) Model diagnostic and check of residuals (FKLM 3.4, 3.5). Lecture notes 25.02.25. Repetition week 8.
Lecture 16 (27.02.2025) Transformations and Inference (FKLM 3.2 3.4, 3.5). Lecture notes 27.02.25.
Lecture 17 (04.03.2025) Prediction, Indicator variables, Orthogonal columns, Restrictions on coefficients. Lecture notes 04.03.25. Repetition week 9.

Lecture 18 (06.03.2025) Multiple hypothesis testing (HBL note). Lecture notes, slides, example in r. Note change of lecturer.
Lecture 19 (11.03.2025) Analysis of variance (HS 8.1.1). Lecture notes, slides, example in r.
Lecture 20 (13.03.2025) Analysis of variance (HS 8.1.1). Lecture notes, slides, example in r.
Lecture 21 (18.03.2025) Design of experiments (DOE), introduction (T note). Lecture notes, slides, DOE example in r.
Lecture 22 (20.03.2025) DOE interactions and blocking (T note). Lecture notes, slides.
Lecture 23 (25.03.2025) DOE fractional factorial designs (T note). Lecture notes, slides, DOE example with plots in r.
Lecture 24 (27.03.2025) Use the lecture time (and room) to work on project and ask questions!
Lecture 25 (01.04.2025) Use the lecture time (and room) to work on project and ask questions!
Lecture 26 (03.04.2025) Repetition part 1, lecture notes.
Lecture 27 (08.04.2025) Repetition part 2; exam 2019 problem 3, exam 2017 problem 2, exam 2014 problem 4.
Lecture 28 (10.04.2025) No lecture, but you can work on repetition and ask questions in R9.

Exam:
Here is the exam questions, and a tentative solution manual. The results are now publised, and here is the grading document

Re-sit exam:
Here is the exam questions, and a tentative solution manual.