Skip to main content
. 2019 Jan 22;20:44. doi: 10.1186/s12859-019-2608-9

Fig. 14.

Fig. 14

The comparison of predictive performances between CDCN I and kNN methods. (a) showing the result on the CCLE dataset with 10% random missing data, and (b) showing the result on the CCLE dataset with 20% random missing data