Research Article

Evolutionary Multilabel Feature Selection Using Promising Feature Subset Generation

Table 4

Comparison results in terms of ranking loss.

MethodBirdsEnronLlogMediamillMedical
Proposed0.115 ± 0.0170.100 ± 0.0080.155 ± 0.0230.060 ± 0.0010.115 ± 0.026
GA0.129 ± 0.0150.133 ± 0.0240.164 ± 0.0230.066 ± 0.0070.145 ± 0.026
NSGA0.125 ± 0.0170.149 ± 0.0310.163 ± 0.0230.067 ± 0.0110.139 ± 0.026
MPSOFS0.132 ± 0.0170.194 ± 0.0110.180 ± 0.0210.159 ± 0.0040.140 ± 0.024

MethodTMC2007BusinessEducationEntertainmentHealth
Proposed0.073 ± 0.0010.062 ± 0.0270.089 ± 0.0030.111 ± 0.0020.085 ± 0.028
GA0.075 ± 0.0020.070 ± 0.0290.100 ± 0.0040.140 ± 0.0070.098 ± 0.027
NSGA0.076 ± 0.0020.067 ± 0.0270.100 ± 0.0030.137 ± 0.0090.097 ± 0.028
MPSOFS0.078 ± 0.0020.096 ± 0.0270.101 ± 0.0030.153 ± 0.0050.098 ± 0.028

MethodReferenceScienceSocialSocietyAvg. rank
Proposed0.111 ± 0.0230.118 ± 0.0030.075 ± 0.0100.135 ± 0.0031.00
GA0.130 ± 0.0210.154 ± 0.0070.088 ± 0.0120.151 ± 0.0082.86
NSGA0.128 ± 0.0250.150 ± 0.0060.085 ± 0.0110.153 ± 0.0112.29
MPSOFS0.140 ± 0.0230.157 ± 0.0040.097 ± 0.0120.212 ± 0.0053.86