Skip to main content
Fig. 2 | BMC Genetics

Fig. 2

From: Data mining and machine learning approaches for the integration of genome-wide association and methylation data: methodology and main conclusions from GAW20

Fig. 2

Recursive feature elimination–random forest applied to combined genome-wide genotype and methylation data. Recursive feature elimination was applied to random forest (RF) and consisted of the following steps: a running the random forest model; b removing features that random forest ranked in the bottom 3%; c ranking removed features starting with the lowest rank; and (d) recursively iterating until no additional features could be removed from the model. The comparison between random forest and random forest with recursive feature elimination relied on the full set of ranks

Back to article page
\