Skip to main content

Table 6 Accuracy of the RF, RFOK, and RFCK models based on the validation data

From: Machine learning and geostatistical approaches for estimating aboveground biomass in Chinese subtropical forests

Year

Models

R2

MAE (t∙ha−1)

RMSE (t∙ha−1)

RMSE (%)

Bias (t∙ha−1)

Bias (%)

RI

1992

RF

0.41

9.93

18.99

112.47

−3.82

−22.59

/

RFOK

0.81

6.64

15.64

92.61

−1.77

−10.48

0.176

RFCK

0.81

6.54

15.62

92.48

−1.80

−10.68

0.177

2002

RF

0.35

34.81

56.67

97.19

−3.31

−5.67

/

RFOK

0.40

33.57

53.81

92.29

−3.19

−5.47

0.050

RFCK

0.41

32.32

53.78

92.24

−2.14

−3.67

0.052

2010

RF

0.49

24.56

31.21

52.08

2.09

−3.89

/

RFOK

0.53

23.47

30.23

52.05

2.26

−3.59

0.031

RFCK

0.55

22.14

29.97

51.60

1.07

−1.85

0.038