Table 3 Mean absolute errors and Pearson correlation coefficients for ML on the dataset of Ravi et al.34, which contains 25 nitropyrazole molecules. 5-fold cross validation was used, so Ntrain = 20 and Ntest = 5.

From: Applying machine learning techniques to predict the properties of energetic materials

  Vdet (km/s) ρ (g/cc) Pdet (GPa)
KRR Estate 0.12, 0.99 0.04, 0.98 199, 0.92
CDS 0.07, 0.99 0.03, 0.99 1.10, 0.99
SoB 0.08, 0.99 0.03, 0.99 0.83, 0.99
Ridge Estate 0.32, 0.91 0.03, 0.98 2.48, 0.98
CDS 0.14, 0.99 0.02, 0.99 1.34, 0.99
SoB 0.44, 0.86 0.03, 0.99 2.92, 0.96
mean n/a 1.25, 0.00 0.27, 0.00 12.90, 0.00