Workshop 4: Prediction

We’ll be looking at using the linear model for predicting a numeric measure. We’ll look at:

• Visualising the relationship between a target variable and other numeric measures.
• Splitting a dataset into training and test/validation.
• Fitting linear models.
• Evaluating the model fit using the test set using (root) mean square error.
• Comparing model predictive performance.

Start by downloading lab04.Rmd and load it into RStudio:

https://www.massey.ac.nz/~jcmarsha/161324/labs/lab04.Rmd

There are a bunch of packages you’ll likely need to install - you’ll probably get them listed in a yellow bar at the top of the editor pane.

Ensure the document knits correctly, then work your way through the R Markdown worksheet.