Standardizing features
It is important to ensure that the feature inputs to the kNN distance calculation are standardized using the scale()
function. Standardization ensures that features with large mean or variance do not disproportionately influence the kNN distance score.
This exercise is part of the course
Introduction to Anomaly Detection in R
Exercise instructions
- Apply the
summary()
function to thewine
data to calculate the mean, minimum and maximum values forpH
andalcohol
. - Use the
scale()
function to create a standardized version of thewine
data calledwine_scaled
. - Use the
summary()
function towine_scaled
to check that the mean and ranges have changed.
Hands-on interactive exercise
Have a go at this exercise by completing this sample code.
# Without standardization, features have different scales
summary(wine)
# Standardize the wine columns
wine_scaled <- ___
# Standardized features have similar means and quartiles
___(___)