Figure 3

The cross-validation results of the PLS model using spectral variables within 1800–900 cm−1. (A) The regression plot between the predicted and actual PMI. The black line represents the reference line where the predicted PMI scores are closer to it, the higher fitting of goodness will be. (B) The plot of VIP scores displays the contribution of the spectral variables to the distinction in the PLS model. The variables with VIP scores above 1.0 (marked by a red dot line) are considered most significant, and their assignments are symbolized.