Variation
Variability: Analysis
Survival Tree
Correlation and Regression
Multiple Regression
Randomized Experiments
You might also read
Articles linked to this work by shared authors, journal, and citation graph.
Kristin K Nicodemus1, James D Malley, Carolin Strobl
1Statistical Genetics, Wellcome Trust Centre for Human Genetics, University of Oxford, Oxford, UK. kristin.nicodemus@well.ox.ac.uk
Random forests (RF) variable importance measures (VIMs) show increased importance for correlated predictors when they are associated with outcomes. Unconditional VIMs are unbiased under the null hypothesis, offering a practical choice for large datasets.
Area of Science:
Background:
Purpose of the Study:
Main Methods:
Main Results:
Conclusions: