Research Article

Comparison of Different EHG Feature Selection Methods for the Detection of Preterm Labor

Table 5

Comparison of the percentage of correct classification of the selected features subset by using LDA.

Selection methodSelected feature subset Correct classification using LDA

JDW3, D6, D7, D8, D9 81.20%
SFS with QDATR, LE, SE, VarEn, W1, W2, W4, D8, D9 84.96%
SFS with LDA SE, VarEn, W2, W3, D1, D3, D7, D8, D9, MPF 83.71%
SFS with KNNTR, LE, SE, VarEn, W1, W2, W3, W4, W5, D1, D2, D3, D4, D7, D8, MPF84.96%
BPSO with QDALE, VarEn, W1, W4, D1, D2, D4, D5, D8, D9 82.46%
BPSO with LDASE, VarEn, W1, W2, D1, D4, D8, MPF 83.71%
BPSO with KNNLE, SE, VarEn, W3, W4, D1, D3, D5, D6, D8 82.46%