Journal of Analytical Methods in Chemistry / 2019 / Article / Tab 4 / Research Article
Simultaneous Recognition of Species, Quality Grades, and Multivariate Calibration of Antioxidant Activities for 12 Famous Green Teas Using Mid- and Near-Infrared Spectroscopy Coupled with Chemometrics Table 4 The predictions of DPPH scavenging activities by OSWLS-SVM.
Prediction samples Actual DPPH scavenging activity (%) Recoveries (%) MIR NIR Fusion data MIR NIR Fusion data 1 36.07 36.07 36.07 97.1 100.7 100.0 2 36.53 36.23 36.23 100.0 100.3 100.0 3 36.07 36.53 35.80 101.6 97.5 103.5 4 36.53 36.07 36.23 94.9 101.3 104.9 5 35.65 36.23 35.65 98.2 103.2 97.3 6 36.16 36.53 36.16 100.0 99.8 100.0 7 36.34 35.65 36.16 100.0 101.8 100.0 8 36.16 36.16 36.16 100.0 100.1 98.2 9 35.65 35.65 35.65 96.8 102.4 99.2 10 36.34 36.16 36.34 101.3 99.1 99.2 11 36.16 47.80 36.16 100.0 100.8 100.0 12 47.92 48.06 47.92 100.0 100.2 100.0 13 48.38 48.38 48.38 100.0 100.2 100.9 14 48.06 48.06 47.92 100.0 99.4 99.6 15 48.38 48.17 48.17 96.5 98.0 100.0 16 40.02 40.02 40.02 103.9 101.6 101.2 17 40.37 40.37 40.37 100.0 97.5 100.0 18 40.32 40.32 40.32 95.8 101.4 98.6 19 40.21 40.21 40.21 105.8 100.3 99.6 20 40.37 40.37 40.37 103.3 96.8 100.0 21 48.67 48.67 48.67 100.0 103.3 103.9 22 49.04 49.04 49.04 95.5 100.6 99.7 23 49.16 48.67 49.16 97.6 100.2 100.0 24 48.81 49.04 48.81 100.0 90.5 100.0 25 49.27 49.16 49.27 100.0 101.6 100.0 26 35.11 35.52 35.11 101.0 98.7 100.0 27 35.52 35.73 35.52 100.0 98.3 96.4 28 35.92 35.92 35.92 100.0 97.6 100.0 29 35.32 35.11 35.32 100.0 99.6 100.0 30 35.73 35.32 35.73 103.5 99.1 101.7 Correlation coefficient 1.0000 0.9904 1.0000 Regression equation y = 0.9998x + 0.0001y = 0.9538x + 0.0173y = 0.9998x + 0.0001Average recoveries (%) 99.8 ± 2.5 99.7 ± 2.4 100.1 ± 1.7 RMSECa 5.7261 × 10−5 0.0122 6.4662 RMSEPb 0.0100 0.0108 0.0065 T(t -test) = 2.045 0.0841 < 0.0820 < 0.0441 <
a RMSEC denotes calibration root-mean-squared error, b RMSEP denotes prediction root-mean-squared error.