Yangyang Zhou

Jun 29, 2023 | mdpi.com | Yangyang Zhou |Zhongke Feng

3.3. Comparison of the Predicted FSV Estimates among the MLR, BP and RF MethodsThe accuracy of the model is validated using 70 sample data from the test set. The estimated values of FSV are compared with measured values in (Figure 4, Figure 5 and Figure 6). For the same set of modeling factors, RF has the best prediction accuracy for FSV, followed by the BP model, and finally the MLR model. This indicates that machine learning methods have an advantage over traditional regression methods for FSV.