Skip to main content
. 2024 Jun 25;14:14590. doi: 10.1038/s41598-024-64386-w

Table 5.

Comparison of the developed ML models.

Metrics Training data Testing data All data
CatB GPR LGBM CatB GPR LGBM CatB GPR LGBM Prop.Eqn (WOR) Prop.Eqn (WWR)
Mean μ 1.005 1.006 1.008 1.026 1.023 1.026 1.01 1.01 1.012 1.003 1.004
CoV 0.062 0.062 0.076 0.141 0.151 0.154 0.085 0.088 0.098 0.207 0.192
R2 0.997 0.997 0.991 0.933 0.947 0.875 0.986 0.988 0.971 0.917 0.937
MAPE 4.41 4.44 5.32 9.32 10.27 9.61 5.39 5.60 6.18 16.80 13.70
RMSE( kN) 36.8 39.6 63.6 160.9 143.3 219.9 79.1 73.2 113.6 115.9 196.7
a20-index 0.994 0.993 0.981 0.899 0.857 0.887 0.975 0.965 0.962 0.68 0.78