Near infrared (NIR) spectroscopy offers a rapid method for estimating many important wood properties, including air-dry density, microfibril angle (MFA) and SilviScan estimated stiffness (EL(SS)). Wood property calibrations may be improved by using non-linear calibration methods. In this study, we compare calibrations developed using partial least squares (PLS) regression and least-squares support vector machine (LS-SVM) regression, a relatively new technique for modelling multivariate, non-linear systems. LS-SVM regression provided the strongest calibration statistics for all wood properties. For an equivalent number of latent variables, the predictive performance of the MFA LS-SVM calibrations were superior to those of the corresponding PLS calibration, while predictive results for air-dry density and EL(SS) were similar for both calibration methods.
BirkettM.D. and GambinoM.J.T., Pap. S. Afr. November/December, 34 (1988).
2.
SchultzT.P. and BurnsD.A., Tappi J.73, 209 (1990).
3.
WrightJ.A.BirkettM.D. and GambinoM.J.T., Tappi J.73, 164 (1990).
4.
GarbuttD.C.F.DonkinM.J. and MeyerJ.H., Pap. S. Afr. April, 45 (1992).
5.
MichellA.J., Appita J.47, 29 (1994).
6.
MichellA.J., Appita J.48, 425 (1995).
7.
OnaT.SonodaT.ItoK.ShibataM.KatoT. and OotakeY., J. Wood Chem. Technol.17, 399 (1997).
8.
OnaT.SonodaT.ItoK.ShibataM.KatoT. and OotakeY., J. Wood Chem. Technol.18, 27 (1998).
9.
OnaT.SonodaT.ItoK.ShibataM.KatoT. and OotakeY., J. Wood Chem. Technol.18, 43 (1998).
10.
RaymondC.A. and SchimleckL.R., Can. J. For. Res.32, 170 (2002).
11.
SchimleckL.R.RaymondC.A.BeadleC.L.DownesG.M.KubeP.D. and FrenchJ., Appita J.53, 458 (2000).
12.
ThygesenL.G., J. Near Infrared Spectrosc.2, 127 (1994).
13.
HoffmeyerP. and PedersenJ.G., Holz Roh Werkst.53, 165 (1995).
14.
SchimleckL.R.MichellA.J.RaymondC.A. and MuneriA., Can. J. For. Res.29, 194 (1999).
15.
ThummA. and MederR., J. Near Infrared Spectrosc.9, 117 (2001).
16.
GindlW. and TeischingerA., J. Near Infrared Spectrosc.9, 255 (2001).
17.
MederR.ThummA. and MarstonD., J. Near Infrared Spectrosc.11, 137 (2003).
18.
EvansR., Holzforschung48, 168 (1994).
19.
EvansR., in Microfibril Angle in Wood: Proceedings of the IAWA / IUFRO International Workshop on the Significance of Microfibril Angle to Wood Quality, Ed by ButterfieldB.G.University of Canterbury Press, Christchurch, New Zealand, p. 116 (1997).
20.
EvansR., Appita J.52, 283 (1999).
21.
SchimleckL.R. and EvansR., Appita J.56, 312 (2003).
22.
SchimleckL.R. and EvansR., IAWA J.23, 217 (2002).
23.
SchimleckL.R. and EvansR., IAWA J.23, 225 (2002).
24.
SchimleckL.R. and EvansR., Holzforschung58, 66 (2004).
25.
WoldS.Kettaneh-WoldN. and SkagerbergB., Chemometr. Intell. Lab. Sys.7, 53 (1989).
26.
VapnikV., in Nonlinear Modelling: Advanced Black-box Techniques, Ed by SuykensJ.A.K. and VandewalleJ.Kluwer Academic Publishers, Boston, USA, p. 55 (1998).
27.
SuykensJ.A.K.Van GestelT.De BrabanterJ.De MoorB. and VandewalleJ., Least-Squares Support Vector Machines.World Scientific, Singapore (2002).
28.
CogdillR.P. and DardenneP., J. Near Infrared Spectrosc.12, 93 (2004).
29.
ShenkJ.S. and WesterhausM.O., Crop Sci.31, 469 (1991).
30.
NæsT.IsakssonT. and KowalskiB., Anal. Chem.62, 664 (1990).
31.
GirosiF.JonesM. and PoggioT., Neural Computation7, 219 (1995).
32.
SchimleckL.R.EvansR. and IlicJ., Can. J. For. Res.31, 1671 (2001).
33.
LucasiusC.B. and KatemanG., Chemometr. Intell. Lab. Sys.19, 1 (1993).
34.
WilliamsP.C. and SoberingD.C., J. Near Infrared Spectrosc.1, 25 (1993).
35.
SchimleckL.R.DoranJ.C. and RimbawantoA., J. Agric. Food Chem.51, 2433 (2003).
36.
SchimleckL.R.MoraC. and DanielsR.F., Can. J. For. Res.33, 2297 (2003).
