Pre-course information

There are few things to do before the course starts. These include both pre-course homework (studying), setting up communication channels and your computer.

Pre-course homework

  • Biostatistics and machine learning is based on mathematics so in order to have a good start at the course we need to revise or pick-up the basics.
  • Go to https://olgadet.github.io/bookdown-mlbiostatistics/ and have a look at the Preliminary Mathematics section, 7 chapters covering basics of mathematical notations, sets, functions, differentiation, integration, vectors and matrices.
  • Find pen and paper and write down the solutions (or your best attempt) to the following exercises:
    • Ex. 1.3: f, g, h
    • Ex. 1.4: e, f, g, h
    • Ex. 2.2
    • Ex. 3.2, 3.3
    • Ex. 4.6: h, i, j
    • Ex. 5.1: i, j
    • Ex. 6.1: e
    • Ex. 7.1: h

Communication

  • We will be using Zoom for discussions and Zulip for chatting in writing
Zoom Meeting
Zulip
  • Join via the invitation link sent via email
  • Upload your photo. Edit your profile. Introduce yourself in the “sayHi” stream
  • And feel free to start the conversation going under “prelim-math” while you’re doing your pre-course homework

What to bring (have)

  • a laptop with R and R-Studio installations (see below)
  • a working web camera and a quiet space to take the course from
  • solutions (or attempts) to the exercises (see above Pre-course homework)
  • a pen and paper to write on
  • a positive attitude

R, R-Studio, R-packages and Rmd

  • During the course we will run scripts locally on laptops using R (programming language) and R-Studio (editor). We will write scripts in using R markdown files (.Rmd), a files that include both an easy-to-write plain text that can also contain chunks of embedded R code.
R

Install R version 3.5.0 or higher

  1. Go to CRAN
  2. Click on the link corresponding to your operating system
  3. Download the recommended files for your system.
  4. Run the installer or move the downloaded files to suitable place on your computer.
Install R Studio
  • Go to the web page rstudio download the installer corresponding to your operating system. Unpack the installer and install the app on a suitable place on your system.
  • You should now be able to fire up R-studio and see something like the following:
  • Note that on some operating systems it will be easier to install and run R and R-studio if you are administrator of your own computer and hence are allowed to install software on your machine. If you do not have these privileges please ask your system administrator to install the latest version of R and R-studio.
R packages
  • By default, R installs a set of packages during installation. R package, is a basic unit of sharable code, that bundles together code, data, documentation and tests. In the course we will be using both default and additional R packages; the latter we need to install. To install a package, we type in console install.packages("package-name")
  • Try installing packages that we will use for decision trees and artificial neural networks

# install decision trees packages
install.packages("rpart")
install.packages("rpart.plot")
install.packages("randomForest")

# install artifical neural network package
install.packages("neuralnet")

  • To use the install package we load them by using library() or require() e.g.

library("neuralnet")
require("randomForest")

  • To see packages index page use help() function

help(package="neuralnet")

Install R packages

To install R packages, open R-Studio and in the console, type

R and R Markdown