exam of Introduction to machine learning

The goal of this exam is perform an analysis on data related to heart disease, in particular, we want to explore the relationship between a target variable - whether patient has a heart disease or not - and several other variables such as cholesterol level, age, …

The data is present in the file 'heartData_simplified.csv', which is a cleaned and simplified version of the UCI heart disease data set

We ask that you explore the data-set and answer the questions in a commented R code (or Rmd if you know how). You should send your code to monique.zahn@sib.swiss by the 7th of August.

Do not hesitate to comment your code to explain to us your thought process and detail your conclusions following the analysis.

description of the columns

• age : Patient age in years
• sex : Patient sex
• chol : Cholesterol level in mg/dl.
• thalach : Maxium heart rate during the stress test
• oldpeak : Decrease of the ST segment during exercise according to the same one on rest.
• ca : Number of main blood vessels coloured by the radioactive dye. The number varies between 0 to 3.
• thal : Results of the blood flow observed via the radioactive dye.
  • fixed -> fixed defect (no blood flow in some part of the heart)
  • normal -> normal blood flow
  • reversible -> reversible defect (a blood flow is observed but it is not normal)
• target : Whether the patient has a heart disease or not

code to read the data

heartData <- read.csv('heartData_simplified.csv')
heartData$target=as.factor(heartData$target)
heartData$sex=as.factor(heartData$sex)
heartData$thal=as.factor(heartData$thal)

questions

1. perform a PCA on the age, chol, thalach, ca and oldpeak features. Do the PCA axes helps you to visually distinguish patients along different categorized features such as target, sex or thal ?

2. perform a Hierarchical Clustering on all features but target. Evaluate the quality of your clustering and explore the different options (distances, clustering method).

3. regression

   1. split your data-set in a train and test set.
   2. use a GLM to model and predict target using the other features with the train set
   3. evaluate your GLM model with the test set.
   4. which features are the most relevant to predict heart disease (target)