Fig. 4: Feature importance plots.
From: Language function following preterm birth: prediction using machine learning
a Importance attributed to each feature by the Boruta algorithm. The first eight features coloured in blue (PSFA, twin status, course of antenatal steroids, any antenatal steroids, sex, PSRD, PSAD, feeding at discharge) are the jointly most predictive features towards the prediction of language outcome. b Importance attributed to features by RF variable importance. c Importance attributed to features by ReliefF expRank. Computation of feature importance depends on the feature selection algorithm used and is expressed relative to the maximum.
