MVE190 / MSG501 Statistical learning with regression models Autumn 26

Jump to today

This page contains the program of the course: lectures, exercise sessions and computer labs. Other information, such as learning outcomes, teachers, literature and examination, are in a separate course PM. The course is in-person for what concerns the lectures (no Zoom, no video-recorded lectures). For compulsory mini-analyses presentations it will be possible to connect via zoom or be in the room (see below).

 

Program

The schedule of the course is in TimeEdit. 

There will be

  • Lectures (Tuesday, Thursday)
  • Q&A + exercises (Tuesdays at 15:15, starting from the third week)
  • 1 computer lab (November 5, 15:15), bring your laptop
  • Compulsory to attend mini-analyses with your presentations

Lectures will be "in-person" in University rooms and will not be recorded.

Notice, there are some activities for which attendance is mandatory either in room or via zoom. These are the Fridays "mini-analyses" (not all Fridays). See the course PM for details and what to do in case you are unable to attend.

 

Lectures (below is a  plan based on last year. Deviations 

week Topics slides/notes code and data files
 MiniAnalysis
45

Tuesday: General intro to the course and mention of several topics: bias in linear regression, least squares; parameters interpretation in simple linear regression;

 

 

 

 

Check Check "lecture 1" in Jörnsten's notes found in Course PM

 

 

no mini analysis this week
45

Thursday: least squares estimates and relation to correlation; interpretation of coefficients with transformed variables. the 5 basic assumptions and residual plots.

 

 

Check "lecture 2" in Jörnsten's notes (linked in the previous lecture).

 

 

 

 

no mini analysis this week
45

Thursday 5 November Computer lab: BRING YOUR LAPTOP with R/RStudio in MVF24 and MVF25: the lab is not structured with presentation of concepts. A TA is there to help you go through some basic exercise for those that are new to R.

 

 

 no mini analysis this week
46

Tuesday lecture: Proof of the unbiasedness of OLS parameter estimators; variance of the estimators; some residuals-based diagnostics; box-cox transformations; 

 

 

 

 
46

Thursday: leverage values; deletion-based diagnostics; MSE and proof of unbiasedness of MSE; unbiasedness of yhat;  expectation and variance of residuals; variability decomposition (SSEr, SStor, SSRegr); Rsquared (coeff of determination); t-test construction

Optional exercises for self-study from Rawling's book: exercise 1.1, exercise 1.4, relevant bits of exercise 1.9, ex. 1.10, ex. 1.16,

 

 

Check "lecture 4" in Jörnsten's notes.

 

(Compulsory) Friday 13 Nov : present  work for mini-analysis 1.

 

 

 
47

Tuesday lecture: completing t-tests; pvalues;   confidence intervals for parameters and for E(Y0). Prediction intervals for Ypred0.

 

Q&A+exercise at 15.15

 

See the relevant bits in lectures 4-5 of Jörsten's notes)

 

no mini analysis this week
47

Thursday: The Simpson's paradox and notation for multiple lin. regression.  Properties of the estimators in multiple regression and sampling distributions. t-test and categorical covariates (not everything); Topics  also found in chapter 9 ("Class variables") in Rawlings et al up to sect. 9.3.

 

Optional exercises for self-study from Rawling's book: exercise 3.5(part a and d), ex. 3.10, ex. 3.11, ex. 3.12 (we never do regression without intercept...but if you are interested...), ex 3.13

 

 

 

 

 

 no mini analysis this week
48

Tue lecture: models with categorical and numerical covariates. 

 

Q&A+exercise at 15.15

 

 

 

 

 
48

Thu:  problems with p-values and large datasets; Multicollinearity;  Variance Inflation Factor; Partial F test;

 

for Partial F: section 2 in Lect 6 of notes;

multicollinearity: pages 100--103 in ISLR and pages 372-373 in Rawlings' et al.
[ISLR= "Intro to Statistical Learning and Regression"]

 

 

(Compulsory) Friday 27 Nov at 13.15: present minianalysis 2

 
49

Tue: greedy variables selection: backward search. Bias/variance tradoff and the pMSE. Training and testing, PMSE with regsubsets.

 

Q&A+exercise at 15.15

 

 

[ISLR: sections 2.2.1-2.2.2, 5.1.1, 6.1.1-6.1.2]

[Rawling's et al: sec 7.3, 7.4]

 

 

no minianalysis this week
49

 

Thu: more on pMSE with categorical covariates; interactions

 

 

 no minianalysis this week

 
50

Tue: ; adj-Rsquared; Kullback-Leibler; AIC, BIC; K-fold CV;

 

Q&A+exercise at 15.15

 

(optional if you are interested) AkaikeEasyIntro.pdf

See also "lecture 8" in Jörnsten's notes.

 

 
50

Thu: LOOCV; hat-matrix; residuals; standardised residuals,  studentised residuals; Cook's distance, DFBETAs; intro GLMs;

 

(if we manage to start GLMs, there is additional support in Agresti's book chapters 4 and 14.4)

 

 

(Compulsory) Friday 11 Dec at 13.15: present minianalysis 3

 
51

Tue: the exponential family; Newton-Raphson; Poisson regression; confidence intervals for GLMs;  asymptotic properties of the MLE; CI for predictions; Wald test; deviance; likelihood ratio test;

 

Q&A+exercise at 15.15 

 

 

 

 no mini analysis this week
51

Thu: Poisson + offset term. Negative binomial regression also with offset.

 

 

 

no mini analysis this week

 

Back to the top

 

Computer lab and software

You are encouraged to attend the lab on Thursday 5 November to experiment with some basic analyses. No further computer lab will be given.

Software: We will use the statistical package R to analyze data, powered via the Rstudio interface. You will need to install both on your computer, see the page Installing R and Rstudio.
No previous knowledge of R is required.

Some useful resources:

If you are familiar to MATLAB or Python, the following may be useful:

 

Back to the top

Course summary:

Course Summary
Date Details Due