Comparing modelling performance of chemometric methods for wood discrimination by near infrared spectroscopy

Comparative wood anatomy is the most accepted (traditional) method for identification. However, there an ongoing search effective where traditional methods may be insufficient in distinguishing on species level. Near-infrared spectroscopy (NIRS) one of developing data Scots pine, black sessile oak and Hungarian were collected examined spectral range 12,000–4000 cm−1 with a resolution 4 cm−1. Data analyzed by partial least squares discriminate analysis (PLS-DA), decision trees (DT), artificial neural networks (ANN) support vector machines (SVM). Raw subjected to multiple scatter correction (MSC), standard normal variate (SNV), Savitzky–Golay derivatives (first [FD], second [SD]) smoothing (Sm) combinations these preprocessing (Sm + FD, Sm SD, FD MSC, SNV). Model performance compared through test accuracies. Accuracies varied between 99–100%, 76–98% 73–96%, genus level, pine species, respectively. PLS-DA SVM found successful models. This study revealed that it possible from near-infrared multivariate analysis.