Figure 2

Scatterplot showing the prediction performance of the PLSLR in the test dataset. Graphs show the PLSLR prediction performance of CH (A) and CRF (B). The dashed line represents the line of unity. If the prediction model was 100% accurate, all data points would fall on this line. The dotted lines indicate the 95% confidence limits of the Normal distribution; these are ±1.24 and ±1.50 for CH and CRF, respectively. Compared with the CH model, predicted values of CRF are close to observed values, indicating that the studied variables can explain a large proportion of the variation in CRF measurements.