Table 5 Performance of different kernel functions in the SVM model for estimating groundwater corrosivity and deposition Indices.

From: AI-based forecasting of groundwater corrosion and scaling indices in semi-arid regions using 25-year data analysis

Kernel

Index

Stage

LSI

RSI

PSI

R2

RMSE

R2

RMSE

R2

RMSE

RBF

Train

0.93

0.1

0.83

0.21

0.72

0.29

Test

0.92

0.11

0.82

0.21

0.69

0.31

Poly.

Train

0.84

0.15

0.73

0.28

0.65

0.33

Test

0.62

0.26

0.61

0.39

0.51

0.45

Linear

Train

0.77

0.18

0.64

0.26

0.55

0.35

Test

0.76

0.20

0.58

0.31

0.52

0.36

  1. Significant values are in bold.
Back to article page