Table 4.

Classification results with Pima Indians Diabetes dataset.

Methodology adopted	Accuracy (%)	Sensitivity (%)	Specificity (%)	Selected features
PCA-ANFIS (10x FC) [17]	89.47	70	71.1	1, 2, 6, 7, 8
LS-ELM (10x FC) [18]	78.21	73.91	80	1, 2, 6, 8
GDA-LS-ELM (10x FC) [18]	79.16	79.1	83.33	1, 2, 6, 7, 8
MLNN with LM (10x FC) [19]	79.62	70	70.31	1, 2, 6, 8
PNN (10x FC) [19]	78.05	71	70.5	2, 6, 8
LDA-MWELM [20]	89.74	83.33	93.75	1, 2, 6, 7, 8
Mean selection method [16]	76.04	71	78	2, 6, 8
Half selection method [16]	75.91	69	79	1, 2, 6, 8
Neural network for threshold selection [16]	76.04	71	78	2, 6, 8
PSO + ELM	91.27	85.26	94.10	1, 2, 6, 8
Proposed SRLPSO + ELM	93.09	91.47	96.29	2, 6, 8