16.7 Independent Component Anysis (ICA)

• Unsupervised
• Finds components that are statistically independent from each other, rather than uncorrelated
• Maximise the ‘non-Gaussianity’ of the ICA components - i.e. non-linear

# Note: ICA requires the "dimRed" and "fastICA" packages.
bean_rec_trained %>%
  step_ica(all_numeric_predictors(), num_comp = 4) %>%
  plot_validation_results() + 
  ggtitle("Independent Component Analysis")

There isn’t much separation between the classes in the first few components, so these independent components don’t separate the bean types.