STK3100 - Introduction to Generalized Linear Models

Lecture plan

Here you find the plan for the coming lectures, as well as an overview of what has been covered in the past lectures. Chapter/section numbers and pages refer to the text book (Alan Agresti: Foundations of Linear and Generalized Linear Models). Read the syllabus list in Leganto for details on exceptions within a chapter.

Plan for coming lectures

Week 46

ML-estimation for GLMMs (sections 9.5.1 and 9.5.2), Generalized Estimation Equations (GEE) for marginal models (sections 9.6.3 and 9.6.4) 

We will further discuss the solution of some exam exercises from previous years

You are highly encouraged to work on these beforehand:

Some will be discussed in the lectures, and some in the group session on Friday.

In week 47 there will be no lectures. On Friday November 24th there will be an "Oracle our" group session, with both Ida and Per August available to answer questions connected to the curriculum.

    Overview of past lectures

    Week 45

    Normal linear mixed models (sections 9.3.1 and 9.3.3)

    Prediction of random effects for normal linear mixed models  (section 9.3.2). Marginal linear models (9.61), marginal and generalized linear mixed models (sections 9.1.2 and 9.1.3), binomial and Poisson GLMMs (sections 9.4.1, 9.4.2, and 9.7)

    Week 44

    Quasi-likelihood methods, variance inflation and model misspecification  (sections 8.1.2 (binomial GLM), 8.2.4 (only to middle of page 276), 8.3.1, 8.3.2, 8.3.3, and 8.3.4)

    Normal linear mixed models (section 9.2, and start 9.3.1)

    Week 43

    GLM models for count data (sections 7.1.1, 7.1.3, 7.1.4, and 7.1.6)

    Overdispersion and negative binomial GLMs  (sections 7.3.1, 7.3.2, 7.3.3, and 7.3.4), zero-inflated GLMs (section 7.4.1), practical illustration (7.5.1, and 7.5.2), Quasi-likelihood methods and variance inflation (sections 8.1.1, 8.1.2 (Poisson GLM), 8.1.3)

    Week 42

    Continue: Baseline-category logit models for multinomial responses (sections 6.1.1-6.1.4, and 6.3.2). Multinomial response models: Ordinal responses, cumulative logit model (sections 6.2.1, 6.2.2, and 6.3.3)

    Thursday, start: GLM models for count data (sections 7.1.1, 7.1.3, 7.1.4, and 7.1.6)

      Week 41

        Variable/model selection (sections 4.6.1, 4.6.2, 4.6.3), and selecting explanatory variables for a GLM: normal linear models and gamma GLMs (section 4.7)

        Link functions for binomial data (sections 5.6.1, 5.6.3, and 5.7.2), summarizing predictive power for GLMs for binomial data (sections 5.2.4 and 5.2.5)

        Week 40

        Week 39

        • Sections 4.3.5 - 4.3.6 and 4.2.5: Confidence intervals and the Delta method, sections 4.4.1 - 4.4.6  and 5.5.1 - 5.5.3: Deviance and Goodness of Fit for GLMs

        Week 38

        • Start on GLMs (more general than the linear model): Section 4.1.3: Canonical link functions, Section 5.1: Link functions for binary data, Section 5.2.1: Interpreting β, and Sections 4.2, 4.5.5 and 5.3.1-5.3.2: Likelihood and large-sample distributions for GLMs
        • Sections 4.3.1-4.3.4: Wald, likelihood-ratio and score tests, and Sections 4.3.5 - 4.3.6 and start 4.2.5: Confidence intervals 
          • Slides (went through 1-4 on Tuesday, 5-11 on Thursday)
          • Smartboard notes from Thursday
          • R code illustrating testing (low birthweight example)

          Week 37

          • Sections 3.2.1 (continue from last week) and 3.2.2
          • Start on GLMs (more general than the linear model): Section 4.1: Exponential family distributions
            • Slides (went through 1-6 on Tuesday and 7-12 on Thursday)
            • R code illustrating grouped and ungrouped binary data (beetles example)
            • Smartboard notes from Thursday

          Week 36

          • Section 2.2: Projections of data onto model spaces (except 2.2.4)
            • Slides (went through 1-14 last week, finished on Tuesday)
            • Smartboard notes (pages 1-4 relevant here) from Tuesday
          • Sections 3.1 (only Cochran's theorem in 3.1.4 and except 3.1.5), started 3.2.1
            • Slides (went through 1-4 Tuesday, Wednesday: 5-12 (did not finish 12, continue next week))
            • Smartboard notes from Tuesday (pages 5-7 relevant here)
            • Smartboard notes from Thursday (the part for slide 12 on page 2 is unfortunately incomplete, continue next week)

          Week 35

          • Section 2.1: Least-squares model fitting
          • Section 2.2: Projections of data onto model spaces (except 2.2.4)

          Week 34

           

          Published July 31, 2023 4:17 PM - Last modified Nov. 15, 2023 2:52 PM
          E-mail this page