Table 3 Case study 2—evaluation results.

From: Proposing a hybrid metaheuristic optimization algorithm and machine learning model for energy use forecast in non-residential buildings

Test data for case 2

Learning WIO-SVR

Testing WIO-SVR

RMSE (kWh)

MAE (kWh)

MAPE (%)

R

RMSE (kWh)

MAE (kWh)

MAPE (%)

R

01/12/2019

1.82

3.33

63.48

0.97

1.04

1.07

4.66

0.89

02/12/2019

1.82

3.30

63.45

0.97

1.64

2.68

7.37

0.95

03/12/2019

1.81

3.29

63.42

0.97

1.58

2.49

9.23

0.96

04/12/2019

1.81

3.28

63.49

0.97

1.63

2.66

10.06

0.95

05/12/2019

1.81

3.28

63.68

0.97

1.40

1.95

7.81

0.97

06/12/2019

1.80

3.25

64.08

0.97

1.33

1.77

6.92

0.96

07/12/2019

1.80

3.22

64.21

0.97

1.27

1.61

6.42

0.95

08/12/2019

1.79

3.20

64.29

0.97

1.01

1.01

6.20

0.87

09/12/2019

1.78

3.17

64.33

0.97

1.31

1.72

6.16

0.95

10/12/2019

1.78

3.18

64.74

0.96

1.22

1.50

5.51

0.97

11/12/2019

1.79

3.20

65.17

0.96

1.42

2.00

6.82

0.95

12/12/2019

1.79

3.22

65.50

0.96

1.21

1.46

5.55

0.96

13/12/2019

1.78

3.18

65.37

0.96

1.31

1.71

5.99

0.96

14/12/2019

1.77

3.15

65.26

0.96

1.51

2.29

7.78

0.94

Average

1.80

3.23

64.32

0.97

1.35

1.85

6.89

0.95

Std. dev

0.02

0.05

0.77

0.00

0.20

0.53

1.47

0.03

Min

1.77

3.15

63.42

0.96

1.01

1.01

4.66

0.87

Max

1.82

3.33

0.65

0.97

1.64

2.68

10.06

0.97

Back to article page