# ini: The initial number of classifiers

# fin: The final number of classifiers

% err: The error percentage (1 - acc)

% acc: The accuracy percentage

l time: The time taken to learn from the training dataset

l time: The time taken to classify the test dataset

% majm: The percentages of misses that were meant to be the majority class

% minm: The percentages of misses that were meant to be the minority class

t missed: The number of environments that could not be classified (no relevant classifier available)

The top line is without IR, and bottom with IR.

================================================================================

# ini # fin % err % acc l time c time maj m min m t missed

================================================================================

Accuracy cutoff: 0.15

--------------------------------------------------------------------------------

500 718.000 0.209 0.791 35.027 12.263 0.008 0.992 0.000

500 664.333 0.191 0.809 33.368 11.953 0.113 0.887 0.000

Accuracy cutoff: 0.3

--------------------------------------------------------------------------------

500 564.333 0.209 0.791 26.829 9.504 0.008 0.992 0.000

500 461.000 0.195 0.805 23.059 8.457 0.098 0.902 0.000

Accuracy cutoff: 0.45

--------------------------------------------------------------------------------

500 502.667 0.209 0.791 27.861 9.615 0.008 0.992 0.000

500 533.333 0.202 0.798 28.781 10.222 0.050 0.950 0.000

Accuracy cutoff: 0.6

--------------------------------------------------------------------------------

500 480.000 0.214 0.786 26.305 9.108 0.001 0.999 0.000

500 499.333 0.207 0.793 27.128 10.225 0.014 0.986 0.000

Accuracy cutoff: 0.75

--------------------------------------------------------------------------------

500 229.667 0.208 0.792 11.813 3.865 0.007 0.993 54.667

500 211.000 0.215 0.785 18.506 3.669 0.020 0.980 87.667