Skip to main content

Table 1 Analysis of several PLS models using full spectra with and without pre-processing methods

From: An efficient method to reduce grain angle influence on NIR spectra for predicting extractives content from heartwood stem cores of Toona. sinensis

Pre-treatment

Calibration

Validation

R2Cal

RMSECal (%)

LVs

R2 v

RMSEV (%)

EC

 No (raw spectra)

0.83

1.36

10

0.64

1.60

 SNV

0.81

1.48

7

0.66

1.68

 1st derivative

0.82

1.38

8

0.47

1.94

 2nd derivative

0.76

1.57

9

0.72

1.58

 SNV+1st derivative

0.83

1.35

9

0.78

1.44

 SNV+2nd derivative

0.79

1.45

8

0.74

1.52

Grain angle

 No (raw spectra)

0.92

11.52

10

0.90

11.76

 SNV

0.98

6.36

15

0.94

10.10

 1st derivative

0.96

8.62

14

0.94

9.26

 2nd derivative

0.98

6.06

16

0.95

9.01

 SNV+1st derivative

0.98

5.43

16

0.95

9.28

 SNV+2nd derivative

0.95

6.23

19

0.94

9.23

  1. R2Cal The coefficient of determination on calibration, RMSECal root-mean-square error on calibration, R2v The coefficient of determination on validation, RMSEV root-mean-square error on validation, LVs latent variables
Back to article page

Plant Methods

ISSN: 1746-4811