Skip to main content

Table 6 Performance metrics of linear models estimating MOE of boles. Italicised values indicate the recommended model; bold values indicate the best value(s) in each column

From: Validation of models predicting modulus of elasticity in Douglas-fir trees, boles, and logs

 

Calibration dataset

Validation dataset

Comments

Fixed effectsa

R 2 marg

R 2 cond

AIC

RMSE

MAPE

TP rate

RMSE

MAPE

TP rate

 
    

(GPa)

(%)

(%)

(GPa)

(%)

(%)

 

V 2

0.45

0.60

432

1.2

9

60

1.3

9

73

 

D

0.14

0.43

493

1.6

12

27

1.1

7

36

 

ρ

0.18

0.43

481

1.5

11

36

1.2

8

27

 

A

0.12

0.16

511

1.5

11

44

1.1

9

36

 

T

0.27

0.53

487

1.5

12

51

1.0

7

46

 

E

0.60

0.73

373

1.0

7

64

1.3

9

91

High TP rate

V 2+ ρ

0.60

0.73

372

1.0

7

64

1.3

9

91

High TP rate

V 2+ H

0.50

0.65

428

1.2

9

64

1.4

8

82

 

D + A

0.31

0.34

484

1.4

10

47

1.2

10

55

 

D + H

0.27

0.53

485

1.5

12

56

1.0

7

46

 

D + ρ

0.31

0.59

444

1.4

10

42

1.2

9

55

 

H + ρ

0.20

0.56

477

1.6

12

33

1.3

10

46

 

T + ρ

0.34

0.54

454

1.3

11

56

1.0

8

36

 

E + T

0.61

0.78

367

1.0

8

64

1.3

9

91

High TP rate

T + A + ρ

0.45

0.50

450

1.2

9

64

1.0

8

64

 

E + D + H

0.61

0.78

368

1.0

8

67

1.3

9

91

High TP rate

V 2 + D + H

0.53

0.76

418

1.3

10

69

1.4

9

82

 

V 2 + ρ + D

0.59

0.75

367

1.0

7

67

1.3

10

82

 

V 2 + ρ + T

0.61

0.77

365

1.0

8

67

1.3

9

91

High TP rate

V 2 + ρ + D + A

0.69

0.74

364

0.9

6

69

1.3

10

73

Best AIC

  1. aAll variables are as defined in Table 2