Feature Importance in R

Feature Importance in R

library(mlbench)
data(BreastCancer)
BC <- BreastCancer

BC <- na.omit(BC)
BC.rf <- randomForest::randomForest(Class ~ ., data = BC, importance=TRUE)

BC.rf$importance

##                      benign   malignant MeanDecreaseAccuracy MeanDecreaseGini
## Id              0.001105289 0.003681141          0.002008497         6.082721
## Cl.thickness    0.046163442 0.025830372          0.039027930        14.527835
## Cell.size       0.096058486 0.076360225          0.088605085        80.750295
## Cell.shape      0.025640011 0.094125483          0.049419857        60.629815
## Marg.adhesion   0.010139762 0.016861759          0.012477486         7.180260
## Epith.c.size    0.014114885 0.015438777          0.014427404        31.348640
## Bare.nuclei     0.082449879 0.056948587          0.073337526        56.125857
## Bl.cromatin     0.014504436 0.033493443          0.021115832        28.863086
## Normal.nucleoli 0.041192903 0.008730619          0.029598163        22.727945
## Mitoses         0.005836512 0.000919002          0.004115768         1.898671

randomForest::varImpPlot(BC.rf, main = "Title: Feature Importance of BreastCancer")