Skip to Main Content
Table 5.

Example for screening data of model built for five wood species stained with three concentrations of iron acetate and tannic acid predicting the response variable L*. PLSR models were built for 14 different transformations (Database) of NIR spectra for the untreated wood species RO, WO, YP, SYP, and WRC. As categorical predictors the three different stain concentrations (0.05, 0.1, 0.2 g/liter) as well as the three different concentrations of tannic acid (0, 300, 600 mg/liter) were used to build the model. The model for transformation, SG7, was selected based on the smallest RMSEPcv and used for further comparison with models built based on other predictor combinations (Fig. 4).

Example for screening data of model built for five wood species stained with three concentrations of iron acetate and tannic acid predicting the response variable L*. PLSR models were built for 14 different transformations (Database) of NIR spectra for the untreated wood species RO, WO, YP, SYP, and WRC. As categorical predictors the three different stain concentrations (0.05, 0.1, 0.2 g/liter) as well as the three different concentrations of tannic acid (0, 300, 600 mg/liter) were used to build the model. The model for transformation, SG7, was selected based on the smallest RMSEPcv and used for further comparison with models built based on other predictor combinations (Fig. 4).
Example for screening data of model built for five wood species stained with three concentrations of iron acetate and tannic acid predicting the response variable L*. PLSR models were built for 14 different transformations (Database) of NIR spectra for the untreated wood species RO, WO, YP, SYP, and WRC. As categorical predictors the three different stain concentrations (0.05, 0.1, 0.2 g/liter) as well as the three different concentrations of tannic acid (0, 300, 600 mg/liter) were used to build the model. The model for transformation, SG7, was selected based on the smallest RMSEPcv and used for further comparison with models built based on other predictor combinations (Fig. 4).
Close Modal

or Create an Account

Close Modal
Close Modal