Computational Intelligence and Neuroscience / 2017 / Article / Tab 2

Research Article

GA-Based Membrane Evolutionary Algorithm for Ensemble Clustering

Table 2

Average performances (in terms of ) over 100 runs by different ensemble clustering methods (the three highest scores of AVE and the three lowest scores of Var in each column are highlighted in bold).

MethodBalanceIrisPima
MAXMINAVEVARMAXMINAVEVARMAXMINAVEVAR

GMEAEC0.7230.6210.6790.001330.9170.8770.8810.000990.8330.7330.8200.00254
CEGA0.6990.5420.6220.005440.9370.7550.9200.007560.7250.6330.6760.00375
CSPA0.7110.5200.6100.009890.9200.7940.8790.004820.8200.7120.7870.00543
HGPA0.6550.5780.6100.000670.8420.7020.8150.000820.8300.6480.7780.01211
MCLA0.6330.4560.5940.010120.8300.7680.7910.001010.8200.6620.7380.00378
KCC0.6940.3770.5440.019820.8780.5440.7420.013510.7350.6980.7160.00012

MethodWineMagic04Seg
MAXMINAVEVARMAXMINAVEVARMAXMINAVEVAR

GMEAEC0.9520.8780.9410.001340.7830.6550.7310.001340.7510.6150.7070.00589
CEGA0.9300.8400.9200.002520.7120.5420.6770.009420.6590.4210.5580.00983
CSPA0.7230.5530.6930.001420.8240.5540.7430.015640.4560.2350.3730.00873
HGPA0.8300.6620.7590.007560.5770.4320.5200.005460.6580.4230.5040.01425
MCLA0.8790.3200.7760.098440.6540.3440.5260.021210.7780.6840.7170.00178
KCC0.8860.2260.7170.112540.7560.4980.6240.008990.7550.5240.6330.00997