Skip to main content
. 2021 May 6;9:e11396. doi: 10.7717/peerj.11396

Figure 2. Feature importance in the BACPHLIP model.

Figure 2

(A) The distribution of importance values for individual features is highly skewed. (B) Cumulative distribution highlighting that the top 20 and top 50 features account for 59% and 85%, respectively, of the overall feature importance values.