TMA4267 Linear Statistical Models
Lectures
| Lectures: | Mondays 10.15-12.00 in S4 |
|---|---|
| Tuesdays 12.15-14.00 in S1 |
First lecture is Monday January 6.
Note that the handouts (slides and class notes) are part of the reading list (pensum). What will you miss if you don't study the handouts? Well, it's mainly part 4 where SSR is defined differently in Bingham and Fry, and the SST decomposition is different, expecially Lecture 15: Partitioning of variance. The derivation of the Principal components, Lecture 25, is only done superficially in ISLR CH10.2.
A pdf-file with 558 pages (DO NOT PRINT!) containing all the 25 lecture slides and available class notes (the same as you find in the table below) is available here: AllHandouts.pdf. If you find misprints (there MUST be), please report to Mette.
Detailed plan with handouts
(for tentative overview scroll down)
NBNB: the reading list is based on the Bingham and Fry (BF) book (chapters 1-4 and 7), James et al (ISLR) chapter 6 and 10.2 and a note on DOE. The slides and class notes can not replace READING the literature on the reading list, but the handouts (class notes and slides) are part of the reading list (pensum).
| Date | Part | Chapter | Topic | Handouts | Exercises in book | Additional activity/notes | |
|---|---|---|---|---|---|---|---|
| 27 | 29.04 | Exam, August 2011, tentative solution | 1up | ||||
| 26 | 28.04 | 1-7 | All | Summing up, exam, final reading list | 1up (including class notes) | ||
| 25 | 31.03 | 7 | ISLRCh6 and ISLRCh10, ISLRslides6, ISLRslides10 | Dimension reduction (ISLR6.3), PCA (ISLR10.2) | 1up (including class notes) | QuizP7.pdf, L25.R, 6.3b(slides 47-57), 10.2a(slides 1-10)10.2b(slides 11-22) | |
| 24 | 25.03 | 7 | ISLRCh6,ISLRslides6 | Regularization (ISLR6.2), Dimension reduction (ISLR6.3) | 1up (including class notes) | L24.R, 6.3a(slides 44-46) | |
| 23 | 24.03 | 7 | ISLRCh6, ISLRslides6 | Regularization (ISLR6.2) | 1up (including class notes) | L23.R, 6.2a (slides 26-32), 6.2b(slides 33-44), 6.2c(slides 45-) | |
| 22 | 18.03 | 7 | ISLRCh6, ISLRslides6 | Model Selection (ISLR6.1) | 1up (class notes), ISLRslides (only slide 1-22) | 6.1a (slides 1-6), 6.1b(slides 7-12), 6.1c(slides 13-16), 6.1d(slides 17-22) | |
| 21 | 17.03 | 6 | DOEnote (pages 20-29) | DOE fractional factorials | 1up (including class notes) | L21.r QuizPart6 | |
| 20 | 11.03 | 6 | DOEnote (pages 15-20) | DOE blocking | 1up (including class notes) | RscriptDOEtreadmill.r | |
| 19 | 10.03 | 6 | DOEnote (pages 1-14) | DOE full 2ink experiments | 1up (including class notes) | limabeans.r | |
| 18 | 04.03 | 5-6 | DOEnote, BF7.2-7.4 | Model transformation and Taylor expansion, Design of experiments | 1up (including class notes) | Rsession2014March5.r | |
| 17 | 03.03 | 5 | BFch4.2+7.1 | ANOVA is MLR, model assessment | 1up (including class notes) | L17.r | |
| 16 | 25.02 | 4 | BFch3.6+not in book | Partial F test, CI and PI. | 1up (including class notes) | ||
| 15 | 24.02 | 4 | BFch 3.4-3.6 | Quiz (4.5+3.1-3.6), Sums-of-squares, F test. | 1up (including class notes) | Quiz | |
| 14 | 18.02 | 4 | BFch 3.4-3.6 | Distribution of SSE and betahat. | 1up (including class notes) | ||
| 13 | 17.02 | 4 | BFch 3.2-3.3 | MLE for sigma, properties of betahat, projection matrices | 1up (including class notes) | L13.racidrain.txt | |
| 12 | 11.02 | 4 | BFch 3.1-3.2 | Multiple regression, normal equations | 1up (including class notes) | ||
| 11 | 10.02 | 3 | BFch 4.4-4.5 | The multivariate normal MLE and conditional mean | 1up (including class notes) | Quiz | |
| 10 | 04.02 | 3 | BFch 4.3-4.4 | Properties of the covariance matrix, the multivariate normal | 1up (including class notes) | ||
| 9 | 03.02 | 3 | BFch 4.3 | Random vectors and matrices, E, Cov, the multivariate normal | 1up (including class notes) | ||
| 8 | 28.01 | 2 | BFch 2.7-2.8 | Two-way ANOVA, random vectors and matrices, E, Cov | 1up (including class notes) | 2.8-2.11 | L8.r L8questions.csv |
| 7 | 27.01 | 2 | BFch 2.6 | One-way ANOVA | 1up (including class notes) | 2.3, 2.5, 2.10 | L7.r |
| 6 | 21.01 | 2 | BFch 2.5-2.6 | Normal sample mean and variance, one-way ANOVA | 1up (including class notes) | 2.6, 2.7 | |
| 5 | 20.01 | 2 | BFch 2.1-2.5 | Chi-square, F, orthogonality, normal sample mean and variance | 1up (including class notes) | 2.1, 2.2 | L5.r |
| 4 | 14.01 | 1 | BFch 1.5-1.7 | Bivariate normal, maximum likelihood, sums of squares | 1up (including class notes and extra mind map) | ||
| 3 | 13.01 | 1 | BFch 1.5-1.7 | Bivariate normal | 1up (including class notes) | L3.r | |
| 2 | 07.01 | 1 | BFch 1.2,1.5 | Galton data, Least squares, bivariate normal | 1up | L2.r | |
| 1 | 06.01 | 1 | BFch 1.1,1.3, 1.4 | Introduction - course and topics, use of software, correlation, Galton data | 1up | Rintro.pdf |
Final lecture plan
| Weeek | Chapter | Topic | Exercise |
|---|---|---|---|
| 2-3 (4L) | BF1 | Part 1: Simple linear regression and the bivariate normal | 1 |
| 4-5 (4L) | BF2 | Part 2: N to Chisq and F distribution, Fishers lemma, Analysis of Variance (ANOVA) | 2, 3 |
| 6-7 (3L) | BF4.3-4.5 | Part 3: Random vectors and matrices and the multivariate normal | 3, 4 |
| 7-9 (5L) | BF3, 4.1-4.2 | Part 4: Multiple linear regression | 4, 5 |
| 10 (2L) | BF7 | Part 5: Model check and transformations | 5 |
| 11-12 (3L) | DOEnote | Part 6: Design of experiments | 6 |
| 12-14 (4L) | ISLRCh6 and ISLRCh10.2(PCA) | Part 7:Model selection, shrinkage and dimensionality reduction | 7 |
| 15 | No lectures due to excursion. But deadline to hand in of compulsory project. | ||
| 16-17 | Easter break | ||
| 18 (2L) | 28.04 10.15-12 in S4 and 29.04 12.15-14 in S1: Summing up and concluding remarks, exam preparation |