Obtain random forest, LDA and KNN test error rates on a test dataset simulated from the same model (for greater stability of the results, use a much larger test set: say, 10K observations). Describe the differences between different methods and across the sample sizes used here