Research Article

Multiresponse Optimization of Pomegranate Peel Extraction by Statistical versus Artificial Intelligence: Predictive Approach for Foodborne Bacterial Pathogen Inactivation

Table 1

Box–Behnken design and ANN models of variables (in coded levels), with phytochemical contents (TPC, TFC, and TAC) and anti-S. aureus and S.enterica activity responses.

Run order(X1)(X2)(X3)Recovery extraction (%)Phytochemical contentAntibacterial activity
TPC (mg GAE/g)TFC (mg QE/g)TAC (mg cy-3-glu/100 g)S. aureus (MIC mg/ml)S. enterica (MIC mg/ml)
Exp.Pred. RSMPred. ANNExp.Pred. RSMPred. ANNExp.Pred. RSMPred. ANNExp.Pred. RSMPred. ANNExp.Pred. RSMPred. ANN

110 (−1)150 (0)10 (−1)18.45 ± 0.92c123.8 ± 6.19cd127.485128.67137.11 ± 1.85f35.24937.064.78 ± 0.22c4.6074.8170.39 ± 0.00b0.340.3950.195 ± 0.00c0.1880.193
210 (−1)180 (+1)20 (0)19.98 ± 0.95d118.02 ± 5.91c115.312118.1530.02 ± 1.45d28.89230.0064.64 ± 0.21c4.8844.8280.39 ± 0.00b0.4920.3820.195 ± 0.00c0.1710.195
360 (+1)150 (0)10 (−1)17.3 ± 0.86b109.6 ± 4.48bc106.637109.84937.09 ± 1.84f35.53536.60611.61 ± 0.52h10.08511.5820.097 ± 0.00a0.2240.1280.195 ± 0.00c0.1850.194
435 (0)150 (0)20 (0)17 ± 0.85b70.02 ± 3.49a70.0071.15622.00 ± 1.22a23.03723.0194.64 ± 0.21c4.6334.5861.56 ± 0.00d1.561.5090.781 ± 0.00b0.7800.772
510 (−1)120 (−1)20 (0)15.6 ± 0.7ab130 ± 6.47d126.162129.78933.35 ± 1.62de34.74933.5073.71 ± 0.17b2.5203.810.39 ± 0.00b0.4640.3860.39 ± 0.00a0.3250.393
635 (0)180 (+1)10 (−1)15 ± 0.75a85.00 ± 4.23ab84.12584.72925.00 ± 1.22b27.86223.0199.75 ± 0.46g9.3859.8080.781 ± 0.00c0.5470.7940.39 ± 0.00a0.4240.388
760 (+1)120 (−1)20 (0)15.15 ± 0.6a99.91 ± 4.92b102.712100.10829.00 ± 1.44cd30.10828.9173.94 ± 0.12b3.4064.0330.39 ± 0.00b0.3470.4110.195 ± 0.00c0.1970.198
835 (0)120 (−1)10 (−1)16.7 ± 0.45b112.5 ± 5.61bc112.65112.1630.01 ± 1.5d30.35729.8977.01 ± 0.33f8.7666.9820.195 ± 0.00e0.1700.2060.195 ± 0.00c0.1960.197
935 (0)150 (0)20 (0)14.4 ± 0.51a70.00 ± 3.41a70.0071.15622.01 ± 1.02a23.03723.0194.61 ± 0.21c4.6334.5861.56 ± 0.00d1.561.5090.781 ± 0.00b0.7800.772
1035 (0)150 (0)20 (0)17.85 ± 0.82bc70.01 ± 3.33a70.0071.15622.00 ± 1.09a23.03723.1114.64 ± 0.22c4.6334.5861.56 ± 0.00d1.561.5090.781 ± 0.00b0.7800.772
1110 (−1)150 (0)30 (+1)17.05 ± 0.79b106.5 ± 5.25bc109.462107.04335.50 ± 1.75e36.96535.7065.32 ± 0.20d5.5474.1120.39 ± 0.00b0.2630.4110.195 ± 0.00c0.1930.193
1260 (+1)150 (0)30 (+1)15 ± 0.66a95.8 ± 4.77b92.11295.75528.00 ± 1.39c29.75128.1516.38 ± 0.30e6.5546.3840.195 ± 0.00e0.2450.2060.39 ± 0.00a0.3640.369
1335 (0)120 (−1)30 (+1)17.27 ± 0.84b73.6 ± 3.66a74.47573.89533.36 ± 1.66de30.50833.0713.02 ± 0.14a3.0683.3410.39 ± 0.00b0.4420.3910.39 ± 0.00a0.4140.415
1435 (0)180 (+1)30 (+1)17.35 ± 0.36b89.9 ± 4.44ab89.7589.39724.00 ± 1.15b23.64324.15414.86 ± 0.71i12.79214.4340.195 ± 0.00e0.2190.1970.39 ± 0.00a0.3810.382
1560 (+1)180 (+1)20 (0)17 ± 0.49b96.6 ± 4.81b100.43796.48228.06 ± 1.32c26.60528.1517.89 ± 0.37f9.4847.9610.781 ± 0.00c0.4750.7970.39 ± 0.00a0.3630.374

Variables: extraction time (X1), agitation speed (X2), and solvent/solid ratio (X3). ±: Standard deviation of three replicates; Averages with different letters (a–i) in the same column, for each parameter, are different ().