Fig. 6
From: Supervised learning methods in modeling of CD4+ T cell heterogeneity
Performance optimization of Random Forest (RF) model. The RF model was created using the randomForest package in R. To optimize the performance of the RF model, two main variables – mtry (numbers of variables randomly sampled as candidates at each split) and ntree (numbers of trees to grow) – were optimized. The RF model with 1000 trees and 4 variables randomly sampled as candidates at each split was identified to perform best