Research Article

Evolutionary Multilabel Feature Selection Using Promising Feature Subset Generation

Table 5

Comparison results in terms of normalized coverage.

MethodBirdsEnronLlogMediamillMedical
Proposed0.194 ± 0.0150.277 ± 0.0150.202 ± 0.0250.196 ± 0.0030.155 ± 0.029
GA0.223 ± 0.0250.337 ± 0.0380.201 ± 0.0210.205 ± 0.0050.180 ± 0.027
NSGA0.212 ± 0.0200.336 ± 0.0380.199 ± 0.0250.208 ± 0.0190.178 ± 0.029
MPSOFS0.222 ± 0.0230.413 ± 0.0140.201 ± 0.0230.330 ± 0.0060.179 ± 0.027

MethodTMC2007BusinessEducationEntertainmentHealth
Proposed0.203 ± 0.0020.132 ± 0.0230.148 ± 0.0040.197 ± 0.0030.157 ± 0.023
GA0.208 ± 0.0040.139 ± 0.0240.159 ± 0.0030.221 ± 0.0090.167 ± 0.026
NSGA0.209 ± 0.0040.141 ± 0.0210.159 ± 0.0040.219 ± 0.0090.167 ± 0.026
MPSOFS0.211 ± 0.0020.168 ± 0.0230.159 ± 0.0040.233 ± 0.0030.166 ± 0.025

MethodReferenceScienceSocialSocietyAvg. rank
Proposed0.157 ± 0.0230.182 ± 0.0040.127 ± 0.0100.242 ± 0.0071.21
GA0.177 ± 0.0220.215 ± 0.0050.140 ± 0.0110.257 ± 0.0062.86
NSGA0.171 ± 0.0210.213 ± 0.0060.136 ± 0.0140.261 ± 0.0082.43
MPSOFS0.179 ± 0.0220.215 ± 0.0060.145 ± 0.0110.317 ± 0.0073.50