Skip to main content

Table 3 Simulation design of predictors for Scenario 3 and 4

From: Intervention in prediction measure: a new approach to assessing variable importance for random forests

Variables X 1 X 2 X 3 X 4 X 5 X 6 X 7
Distribution B(1,0.6) N(3,1) Unif(4,6) N(5,1) X 2 + N(0,0.15) P(X 2) P(X 1)
  1. The variables are sampled independently from the following distributions. B(1, π) stands for the Binomial distribution with n = 1, i.e the Bernoulli distribution, and probability π. N(μ,σ) stands for the normal distribution with mean μ and standard deviation σ. Unif(a,b) stands for the continuous uniform distribution on the interval [a,b]. P(X) stands for random permutation of the values generated in the variable X