Figure 5

The effect of the number of components on the predictive ability of a PLS model. The red dashed line is a model based on all of the features, whereas the model represented by the blue solid line is based only on the subset selected by the WAAC algorithm. The best subset line ends at 59 components, as there are only 59 features in this subset. The line for all features is truncated at 174 components as the RMSE rapidly increases after this point.