Correlated variables
In this exercise, you will inspect the dataset with respect to correlated variables. It is important to remove them before applying a binary classifier, especially in the case of logistic regression. When two or more variables are highly correlated you should remove all except for one.
First, we will use the corrplot() function in the corrplot package to visualize the correlations.
In the correlation plot, blue represents a positive correlation and red a negative correlation.
A darker color indicates a higher correlation.
Finally, you will remove the highly correlated variables from the data set.
This exercise is part of the course
Predictive Analytics using Networked Data in R
Hands-on interactive exercise
Have a go at this exercise by completing this sample code.
# Remove the Future column from studentnetworkdata
no_future <- ___
# Load the corrplot package
library(___)
# Generate the correlation matrix
M <- ___(no_future)
# Plot the correlations
___(M, method = "circle")