Linear Models in R Workshop

Event Name Linear Models in R Workshop
Start Date 25 Oct 2018 11:00 AM
End Date 25 Oct 2018 4:00 PM
Duration 5 hours
Description

DESCRIPTION

This is an introduction to using linear models in R. Linear models are of very general use in statistics, and an important building block in R. Applications include fitting lines and curves to data, making predictions based on a set of predictor variables, performing t tests and analysis of variance tests and calculating confidence intervals on quantities of interest. The mechanics of working with linear models in R is foundational to further statistical and machine learning methods including, as a very incomplete list, logistic regression, survival analysis, mixed models, and differential gene expression analysis.

This 3 hour workshop will consist of a modest amount of theoretical presentation, emphasising geometric intuition rather than maths, and hands on use of R.

Prerequisites:
Attendees will need some existing exposure to R, for example from one of our previous "Introduction to R" workshops. Attendees should bring a laptop. The only required software is a web browser.

Attendees will learn how to:

  • specify terms of a linear model that can serve as a model of their experimental design, such as continuous and categorical independent variables, interactions, and batch effects.
  • fit a linear model to data, interpret the meaning of coefficients, and use the model to make predictions.
  • diagnose problems fitting a linear model.
  • test a hypothesis using two nested linear models. Many common statistical tests can be performed in this way.
  • calculate contrasts, test hypotheses involving contrasts, and calculate confidence intervals for contrasts.
  • as an example application of particular interest in bioinformatics, construct design matrices and contrasts for RNA-Seq differential gene expression analysis with complex experimental designs.

Location: E365 Lecture Theatre, 20 Chancellors Walk, Clayton Campus

Contact: datafluency@monash.edu