Table 6 Summary of statistical evaluation metrics for statistical baseline models, where bold highlighted for the best models.

From: Salt rock creep deformation forecasting using deep neural networks and analytical models for subsurface energy storage applications

Sr. no.

TS datasets (MPa)

Classical or baseline models

MAPE (%)

SMAPE (%)

RMSE

MAE

1

5

ES

16.04

14.49

3.72

3.19

TBATS

17.22

15.48

3.95

3.45

Auto ARIMA

17.36

15.59

3.98

3.45

Theta

2.42

2.42

0.55

0.48

2

15

ES

7.56

7.85

0.92

0.77

TBATS

5.25

5.16

0.59

0.51

Auto ARIMA

5.82

5.88

0.68

0.58

Theta

6.35

6.06

0.74

0.60

3

25

ES

32.55

26.90

8.16

7.06

TBATS

2.66

2.65

0.72

0.57

Auto ARIMA

8.30

7.88

2.04

1.79

Theta

2.18

2.18

0.53

0.47

4

35

ES

2.93

2.99

0.40

0.32

TBATS

20.04

22.66

2.39

2.21

Auto ARIMA

4.85

4.71

0.68

0.53

Theta

7.31

7.61

0.85

0.80