TABLE 3.
Performance and optimum number of features of IFS with PART when using different feature ranking methods.
Feature ranking | Number of optimum features | Number of classification rules | MCC | Overall accuracy |
mRMR | 200 | 7085 | 0.709 | 0.730 |
MCFS | 400 | 7413 | 0.781 | 0.798 |