BIOS601 AGENDA: MTW Oct 02/03/04 and 16/17/18, 2023

AGENDA: MTW Oct 02/03/04, 2023

The (log-)likelihood, and parameter estimation (C&H Ch 03): scalar and vector parameters

• Material: C&Hs Chapter 03 (Likelihood)
  
  
• Exercises to be written up in a joint Markdown document, after we have gone through them in the class/lab
  
  Tentative schedule
  
  
  • MONDAY Oct 2
    
       3.6 Estimate β from 2 datapoints and a model
    
       3.1 Estimate μ from 3 datapoints and Daniel Bernoulli's simple 1-parameter error model
    
    
  • TUESDAY Oct 3
    
       3.2 : Estimate σ of N(0,σ) model from Fisher's binned data
    
    
  • WEDNESDAY Oct 4
    
       3.3 : Estimate μ and σ of N(μ,σ) model from Galton's binned data
    
       3.10 Estimate concentration μ from Fisher's dilution series
    

AGENDA: MTW Oct 16/17/18, 2023 (see *** Footnotes)

• Further examples from Ch 03;
  
     3.4 Did Isaac Newton know the √n law? [PhD, individually.]
  
     3.7 Tumbler Mortality: parameters fitted in 2 ways [team]
  
     3.14 Age at Menarche in 1965 [team]
  
     3.15 Construct an exercise based on 'Timing of puberty ...' article (21st century data) [individually]
  
     3.21 Luck And Skill In Tournament Golf [optional^]
  
  
• Material: C&H Chapter 05 (Rates)
  
     5.01 1-parameter model for rate in control arm, Uganda trial of adult circumcision [individually]
  
     5.03 3-parameter model control arm, Uganda trial [individually]
  
     5.08 Constant-rate model fitted to data from control arm, Kenya trial [individually]
  
  

  *** Footnotes

  
  Team submissions should include an .Rmd file with embedded code,
  so that JH can compile it into a .html file -- and then go back and
  add feedback comments to the .Rmd file, and recompile.
  
  .Rmd files should include verbal explanations/narratives that explain
  the thinking/priciples behind the approaches taken, and the code chunks
  should contain comments explaining the code.
  
  They should also be a statement as to who contributed what to the report
  [academics must do this when submitting manuscripts to a journal].
  
  For Q.15, also provide model answers.
  It would help if the exercise had some 'data' to make it more realistic.
  See instructions given when he used this strategy with course EPIB607 and EPIB693/4.
  
  ^ JH is happy to join whoever wants to to make 1 overall team to
  produce a single .Rmd file with answers to this exercise.