Skip to main content
. 2012 Dec 12;6(Suppl 2):S6. doi: 10.1186/1752-0509-6-S2-S6

Figure 1.

Figure 1

The importance of all 68 features (dataset1). Feature importance generated by random forests. The top-10 features were picked out and various combinations were tested by 10-fold cross-validation to find the best feature subset for prediction of hot spots.