Close
  Indian J Med Microbiol
 

Figure 7: Performance of random forests trained and tested in a 10-fold cross-validation experiment, using combined data from 2012 and 2013 (and an extended feature set). Algorithm used: Random Forest from Weka: 200 trees, max depth 5; cost matrix applied for unbalanced set

Figure 7: Performance of random forests trained and tested in a 10-fold cross-validation experiment, using combined data from 2012 and 2013 (and an extended feature set). Algorithm used: Random Forest from Weka: 200 trees, max depth 5; cost matrix applied for unbalanced set